ailia-models

ailia-models

跨平台AI推理SDK,支持多种编程语言和模型

ailia SDK 是一款跨平台高速AI推理工具,适用于Windows、Mac、Linux、iOS、Android、Jetson和Raspberry Pi平台,并支持Unity (C#)、Python、Rust、Flutter (Dart) 和 JNI。该SDK通过Vulkan和Metal利用GPU增强计算性能。用户可通过pip3安装并使用丰富的预训练模型库,包括动作识别、异常检测和音频处理等领域。最新更新增加了mahalanobis-ad和t5_base_japanese_ner模型。

ailia SDKAI模型跨平台GPU加速高性能推理Github开源项目
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/b202433b-dee5-4123-a928-d3b9c26607ec.png">

预训练的先进AI模型集合。

关于ailia SDK

ailia SDK 是一个自包含、跨平台、高速推理的AI SDK。ailia SDK在Windows、Mac、Linux、iOS、Android、Jetson和Raspberry Pi平台上提供统一的C++ API。它还支持Unity(C#)、Python、Rust、Flutter(Dart)和JNI,以实现高效的AI应用。ailia SDK通过Vulkan和Metal广泛利用GPU,实现加速计算。

如何使用

新功能 - ailia SDK现在可以通过"pip3 install ailia"安装!

ailia MODELS教程

ailia MODELS教程 日文版

支持的模型

截至2024年8月9日有340个模型

最新更新

  • 2024.08.09 添加mahalanobis-ad, t5_base_japanese_ner
  • 2024.08.08 添加sdxl-turbo, sd-turbo
  • 2024.08.05 从Transformers迁移到ailia Tokenizer 1.3
  • 2024.07.16 添加grounded_sam
  • 2024.07.12 添加llava
  • 2024.07.09 添加GroundingDINO
  • 更多信息请见我们的Wiki

动作识别

模型参考导出自支持的Ailia版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/f03f7f31-f353-4709-a117-ebb9cfe95e8d.jpg" width=128px>marsMARS: Motion-Augmented RGB Stream for Action RecognitionPytorch1.2.4及以上EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/7ef31810-6b61-47cf-9a32-1c86720200a5.png" width=128px>st-gcnST-GCNPytorch1.2.5及以上EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/7e0aeaa2-bd6c-4456-b4d1-b0a6609e86de.gif" width=128px>ax_action_recognitionRealtime-Action-RecognitionPytorch1.2.7及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/18971153-65c3-4aa2-801f-cf8a17d5a7c7.png" width=128px>va-cnnView Adaptive Neural Networks (VA) for Skeleton-based Human Action RecognitionPytorch1.2.7及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/27f48ffc-dc88-4ee5-bb83-f450c119fd70.png" width=128px>driver-action-recognition-adasdriver-action-recognition-adas-0002OpenVINO1.2.5及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/9292f11b-3ea8-4bad-88c2-62aff9f747e8.gif" width=128px>action_clipActionCLIPPytorch1.2.7及以上

异常检测

模型参考导出自支持的Ailia版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/e889ecf5-75ac-462c-b457-15d26ee3ee16.png" width=128px>padimPaDiM-Anomaly-Detection-Localization-masterPytorch1.2.6及以上EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/37713ecf-74a6-4973-aa9f-5bf5109a420e.png" width=128px>spade-pytorchSub-Image Anomaly Detection with Deep Pyramid CorrespondencesPytorch1.2.6及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/98587d4f-868e-402d-87f5-22ccdb05c5b0.png" width=128px>patchcorePatchCore_anomaly_detectionPytorch1.2.6及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/f190e378-4740-4277-8076-a5ecdd302d81.png" width=128px>mahalanobisadMahalanobisAD-pytorchPytorch1.2.9及以上

音频处理

模型参考导出自支持的Ailia版本博客
crnn_audio_classificationcrnn-audio-classificationPytorch1.2.5及以上EN JP
deepspeech2deepspeech.pytorchPytorch1.2.2及以上EN JP
pytorch-dc-ttsEfficiently Trainable Text-to-Speech System Based on Deep Convolutional Networks with Guided AttentionPytorch1.2.6及以上EN JP
unet_source_separationsource_separationPytorch1.2.6及以上EN JP
transformer-cnn-emotion-recognitionCombining Spatial and Temporal Feature Representions of Speech Emotion by Parallelizing CNNs and Transformer-EncodersPytorch1.2.5及以上
auto_speechAutoSpeech: Neural Architecture Search for Speaker RecognitionPytorch1.2.5及以上EN JP
voicefilterVoiceFilterPytorch1.2.7及以上EN JP
whisperWhisperPytorch1.2.10及以上JP
clapCLAPPytorch1.2.6及以上
wespeakerWeSpeakerOnnxruntime1.2.9及以上
tacotron2Tacotron2Pytorch1.2.15及以上JP
silero-vadSilero VADPytorch1.2.15及以上JP
rvcRetrieval-based-Voice-Conversion-WebUIPytorch1.2.12及以上JP
crepetorchcrepePytorch1.2.10及以上JP
vall-e-xVALL-E-XPytorch1.2.15及以上
hifiganHiFi-GANPytorch1.2.9及以上
distil-whisperHugging Face - Distil-WhisperPytorch1.2.16及以上
microsoft clapCLAPPytorch1.2.11及以上
narabasnarabas: Japanese phoneme forced alignment toolPytorch1.2.11及以上
rnnoisernnoiseKeras1.2.15及以上
audioset_tagging_cnnPANNs: Large-Scale Pretrained Audio Neural Networks for Audio Pattern RecognitionPytorch1.2.9及以上
deep music enhancerOn Filter Generalization for Music Bandwidth Extension Using Deep Neural NetworksPytorch1.2.6及以上
pyannote-audioPyannote-audioPytorch1.2.15及以上JP
kotoba-whisperkotoba-whisperPytorch1.2.16及以上
reazon_speechReazonSpeechPytorch1.4.0及以上版本
reazon_speech2ReazonSpeech2Pytorch1.4.0及以上版本
gpt-sovitsGPT-SoVITSPytorch1.4.0及以上版本JP

背景移除

模型参考导出自支持Ailia版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/82bd96f8-2c38-4a8f-831e-f017e12e6107.png" width=128px>U-2-NetU^2-Net: Going Deeper with Nested U-Structure for Salient Object DetectionPytorch1.2.2及以上EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/196db520-10b8-4ebc-a7ec-a4c98e8515a1.png" width=128px>u2net-portrait-mattingU^2-Net - Portrait mattingPytorch1.2.7及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/b84b93cd-6df3-4094-bce4-8c456fe92e6a.png" width=128px>u2net-human-segU^2-Net - human segmentationPytorch1.2.4及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/55995167-4026-4b74-a195-67c59ff387a1.png" width=128px>deep-image-mattingDeep Image MattingKeras1.2.3及以上EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/2fc346fc-5b6b-4f30-ba13-cbd3a4b2aae2.png" width=128px>indexnetIndices Matter: Learning to Index for Deep Image MattingPytorch1.2.7及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/ce902f19-1bcf-4e88-a8e7-53b040504243.jpg" width=128px>modnetMODNet: Trimap-Free Portrait Matting in Real TimePytorch1.2.7及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/2634f88e-e4e3-4fdb-8794-036f25f4b1fc.png" width=128px>background_matting_v2Real-Time High-Resolution Background MattingPytorch1.2.9及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/20c31422-d373-4d9e-b503-1d4a2044299d.png" width=128px>cascade_pspCascadePSPPytorch1.2.9及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/1044eadd-d9c1-404a-a15b-960742d65815.png" width=128px>rembgRembgPytorch1.2.4及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/17a7afb6-ccd6-4534-a966-826371561027.jpg" width=128px>dis_segHighly Accurate Dichotomous Image SegmentationPytorch1.2.10及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/3b98f518-6bec-47f6-a539-335aef4b8bca.png" width=128px>gfmBridging Composite and Real: Towards End-to-end Deep Image MattingPytorch1.2.10及以上

人群计数

模型参考导出自支持Ailia版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/8ee302d5-35a9-4a21-9616-ab0758f078dd.png" width=256px>crowdcount-cascaded-mtl基于CNN的层叠多任务学习的<br/>高层先验与密度估计的人群计数<br/>(单图像人群计数)Pytorch1.2.1及以上EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/44f3a104-6277-4fc9-9346-77d6a09ee886.png" width=256px>c-3-framework人群计数代码框架(C^3-Framework)Pytorch1.2.5及以上

深度时尚

模型参考来源支持的 Ailia 版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/4765886a-f66e-48e9-8912-e988e7e3aa0d.png" width=128px>clothing-detectionClothing-DetectionPytorch1.2.1 及更高版本EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/42bc7b53-a7b7-4037-a634-6c4b4205029d.png" width=128px>mmfashionMMFashionPytorch1.2.5 及更高版本EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/a8181cd0-64ad-429e-9163-7535463f7f1a.png" width=128px>mmfashion_tryonMMFashion virtual try-onPytorch1.2.8 及更高版本
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/047fa535-6d3d-40d0-b865-de8dd7a574d5.png" width=128px>mmfashion_retrievalMMFashion In-Shop Clothes RetrievalPytorch1.2.5 及更高版本
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/8b833680-14d9-4e35-9d90-f6a1189bde64.png" width=128px>fashionai-key-points-detectionA Pytorch Implementation of Cascaded Pyramid Network for FashionAI Key Points DetectionPytorch1.2.5 及更高版本
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/c86d106f-b675-4632-80fb-03d7b9bc90b1.jpg" width=128px>person-attributes-recognition-crossroadperson-attributes-recognition-crossroad-0230Pytorch1.2.10 及更高版本

深度估计

模型参考来源支持的 Ailia 版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/98bcc875-9bea-4359-b8d4-eb9e59e38852.png" width=256px>monodepth2Monocular depth estimation from a single imagePytorch1.2.2 及更高版本
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/591f00ff-d600-4118-bf61-58d2cf1ed943.png" width=256px>midasTowards Robust Monocular Depth Estimation:<br/> Mixing Datasets for Zero-shot Cross-dataset TransferPytorch1.2.4 及更高版本EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/4fb2812e-5aff-4938-9ec7-fff72de229fa.png" width=256px>fcrn-depthpredictionDeeper Depth Prediction with Fully Convolutional Residual NetworksTensorFlow1.2.6 及更高版本
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/4c910392-e7ed-4086-be74-bf9cc67fa180.png" width=256px>fast-depthICRA 2019 "FastDepth: Fast Monocular Depth Estimation on Embedded Systems"Pytorch1.2.5 及更高版本
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/05215b12-b19d-45d5-9ea6-1f2f28affcbf.png" width=256px>lap-depthLapDepth-releasePytorch1.2.9 及更高版本
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/29520500-ded6-416d-b4cf-5a5a0df5e938.png" width=256px>hitnetONNX-HITNET-Stereo-Depth-estimationPytorch1.2.9 及更高版本
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/bfd13b88-a292-4274-863c-cc9614eb0e03.png" width=256px>crestereoONNX-CREStereo-Depth-EstimationPytorch1.2.13 及更高版本
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/2358298d-a83a-4234-a605-aea0c07f3de3.png" width=256px>mobilestereonetMobileStereoNetPytorch1.2.13 及更高版本
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/6671ea8c-81d9-47ca-8e7d-301ba1845913.png" width=256px>zoe_depthZoeDepthPytorch1.3.0 及更高版本
模型参考来源支持的Ailia版本博客
:-----------------------::------------::------------::------------::------------:
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/aff53b34-f25a-42bb-ba1b-b7adbe517ac2.png" width=128px>latent-diffusion-txt2imgLatent Diffusion - txt2imgPytorch1.2.10及以后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/bfe1f14c-3fdd-4eea-8595-2bf159a8e494.png" width=128px>latent-diffusion-inpaintingLatent Diffusion - inpaintingPytorch1.2.10及以后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/f649f657-4dfa-4885-97e9-403348de4922.png" width=128px>latent-diffusion-superresolutionLatent Diffusion - Super-resolutionPytorch1.2.10及以后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/9f950869-99ca-464a-b967-2ba5cbada35d.png" width=128px>stable-diffusion-txt2imgStable DiffusionPytorch1.2.14及以后JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/f73d04d7-beb8-47e3-bcd6-fd07d5e8c1bb.png" width=128px>control_netControlNetPytorch1.2.15及以后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/b5c38ff8-8a17-47d2-89b6-ecdc6fd5c72d.png" width=128px>DA-CLIPDA-CLIPPytorch1.2.16及以后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/23afa6d2-88aa-44db-b6fb-7f2909b8cae5.png" width=128px>riffusionRiffusionPytorch1.2.16及以后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/9db16dc0-b80c-487b-91d0-ae4768c0c712.png" width=128px>marigoldMarigold: Repurposing Diffusion-Based Image Generators for Monocular Depth EstimationPytorch1.2.16及以后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/7aea27e6-45d3-45d9-9905-bb656ce2e995.png" width=128px>sdxl-turboHugging Face - SDXL-TurboPytorch1.2.16及以后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/8c9d52e0-7723-46c1-a383-50f209b00d31.png" width=128px>sd-turboHugging Face - SD-TurboPytorch1.2.16及以后

面部检测

模型参考来源支持的Ailia版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/3adc6145-9773-4e69-aaf0-6eb79a04e0f3.png" width=128px>yolov1-faceYOLO-Face-detectionDarknet1.1.0及以后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/77f3e1e3-5008-4fc6-a4ef-c7ef572c751f.png" width=128px>yolov3-faceFace detection using keras-yolov3Keras1.2.1及以后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/1e45564f-0082-4fa4-9a4d-b0328a097868.png" width=128px>blazefaceBlazeFace-PyTorchPytorch1.2.1及以后EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/d0c0793a-58c4-4680-9384-4e19ac0b187d.png" width=128px>face-mask-detectionFace detection using keras-yolov3Keras1.2.1及以后EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/8e8f2cc5-e7a1-4b10-8dc3-d43c8d024951.png" width=128px>dbfaceDBFace : real-time, single-stage detector for face detection, <br/>with faster speed and higher accuracyPytorch1.2.2及以后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/c5a754ce-7016-4c09-8637-96aef6ae39fa.png" width=128px>retinafaceRetinaFace: Single-stage Dense Face Localisation in the Wild.Pytorch1.2.5及以后JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/0ccda8e8-9633-44f5-a9a7-29a8aa061d3d.png" width=128px>anime-face-detectorAnime Face DetectorPytorch1.2.6及以后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/51395b79-2e1a-4922-9c52-2aae81880cf1.png" width=128px>face-detection-adasface-detection-adas-0001OpenVINO1.2.5及以后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/3fe468b0-3833-44a0-aaf7-5201783c1efe.jpg" width=128px>mtcnnmtcnnKeras1.2.10及以后
模型引用导出自支持的Ailia版本博客
:-----------------------::------------::------------::------------::------------:
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/0141d187-5d3c-4f6c-aaf4-5f76f540e753.jpg">vggface2用于人脸识别的VGGFace2数据集Caffe1.1.0及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/dd1c1553-826e-4443-b713-34a1fd9acc57.jpg" width=128px>arcfaceArcFace的Pytorch实现Pytorch1.2.1及以上英文 日文
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/ae3512ef-c095-4adc-a025-595d0fd935bd.png" width=128px>insightfaceInsightFace:2D和3D人脸分析项目Pytorch1.2.5及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/935ea670-af41-4c00-838c-d1ecbae91897.jpg" width=128px>cosfaceCosFace的Pytorch实现Pytorch1.2.10及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/dd41dfad-6d3d-4b2d-9363-b47f07a6a3bb.jpg" width=128px>facenet_pytorch使用Pytorch进行人脸识别Pytorch1.2.6及以上

人脸识别

模型参考导出自支持的 Ailia 版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/efde1973-7e1d-4701-8df8-9c0d7841f098.png">face_classification实时面部检测和情感/性别分类Keras1.1.0 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/53b0bc92-1c6b-4dc0-af20-1c57ad16b225.png" width=128px>facial_featurekaggle-facial-keypointsPytorch1.2.0 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/bb6d1110-d27a-4c1e-b36e-7a02620c6d11.png" width=128px>face_alignment使用 Pytorch 构建的 2D 和 3D 面部对齐库Pytorch1.2.1 及以上EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/6c8e10b5-a88b-4ffc-ad75-1bcf2ba77474.png" width=128px>prnet联合 3D 面部重建和密集对齐 <br/>使用位置图回归网络TensorFlow1.2.2 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/28fc3b7c-2370-44ba-a223-05fe41cffee1.png" width=128px>gazeml基于 Tensorflow 的深度学习框架 <br/>用于训练高性能的注视估计TensorFlow1.2.0 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/5c6e2f56-8dcf-456e-88ef-01fa79f1bce4.png" width=128px>facemeshfacemesh.pytorchPytorch1.2.2 及以上EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/612192e5-021e-4bed-8128-1cce425683b1.png" width=128px>mediapipe_irisirislandmarks.pytorchPytorch1.2.2 及以上EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/330b106e-5ca6-4fa6-9720-3dc2f023ec5f.png" width=128px>hopenetdeep-head-posePytorch1.2.2 及以上EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/3b6f4bbb-d3ef-420c-b8b2-24a94e992acf.png" width=128px>ax_gaze_estimationax 视线估计Pytorch1.2.2 及以上EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/f839d1fa-a6cf-45b9-818f-7f0dd3511849.jpg" width=128px>age-gender-recognition-retailage-gender-recognition-retail-0013OpenVINO1.2.5 及以上EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/0f17b352-6733-4317-a1c0-9e710a8b10aa.png" width=128px>ferplusFER+CNTK1.2.2 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/55f24509-ffcf-46e1-b6bb-6a149c165345.png" width=128px>face-anti-spoofing轻量化的面部防伪Pytorch1.2.5 及以上EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/3c36803e-8462-4d90-9061-e88767970bd6.jpg" width=128px>ax_facial_featuresax 面部特征Pytorch1.2.5 及以上EN
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/884df0f4-74a0-4d05-92d5-678044ef65be.png" width=128px>6d_repnet6D 旋转表示用于无约束头部姿势估计 (Pytorch)Pytorch1.2.6 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/ad59d259-c3dd-4b28-bd06-70e515616e4d.png">hsemotionHSEmotion(高速面部情感识别)库Pytorch1.2.5 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/58c3510b-312c-4308-b029-e76424e300cb.png" width=128px>facemesh_v2MediaPipe 面部关键点检测Pytorch1.2.9 及以上版本JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/a2a16e80-5a90-4e2e-acc8-63a0ecd9b40c.png" width=128px>3ddfaTowards Fast, Accurate and Stable 3D Dense Face AlignmentPytorch1.2.10 及以上版本
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/82e32934-2385-4d7c-aa66-0ff7aed06380.png" width=128px>mivoloMiVOLO: 多输入Transformer 用于年龄和性别估计Pytorch1.2.13 及以上版本
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/9c714085-d19b-4386-80ce-070b1706e11c.png" width=128px>L2CS_NetL2CS_NetPytorch1.2.9 及以上版本

面部交换

模型参考导出自支持的 Ailia 版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/e62492c1-bdc1-4385-b360-f5b676243337.png" width=128px>facefusionFaceFusionONNXRuntime1.2.10 及以上

帧插值

模型参考导出自支持的 Ailia 版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/5540df40-f26c-47ef-982b-16e6f5e7a711.png" width=128px>flavrFLAVR: Flow-Agnostic Video Representations for Fast Frame InterpolationPytorch1.2.7 及以上EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/77a7c30f-b95b-4679-9f7d-5cfbba6213be.png" width=128px>cainChannel Attention Is All You Need for Video Frame InterpolationPytorch1.2.5 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/183a738b-eb13-4862-b26f-e6d0b59e3538.png" width=128px>filmFILM: Frame Interpolation for Large MotionTensorflow1.2.10 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/9e9f4f4a-c749-4ef0-8dfd-21cc0155a1dd.png" width=128px>rifeReal-Time Intermediate Flow Estimation for Video Frame InterpolationPytorch1.2.13 及以上

生成对抗网络

模型参考导出自支持的 Ailia 版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/2c413cb1-ece4-43be-9eff-f2b0937dfcac.png">pytorch-ganPytorch GAN Zoo 项目的代码库 (用于训练此模型)Pytorch1.2.4 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/ba6f5ff8-9af7-4bea-af55-a2730750a108.png" width=128px>council-ganCouncil-GANPytorch1.2.4 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/0b90e4be-aaeb-4dda-b67f-a2f2f13e8942.jpg" width=128px>restyle-encoderReStylePytorch1.2.9 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/8e6464ac-d0d2-46e8-a805-39f947b8be07.jpg" width=128px>samAge Transformation Using a Style-Based Regression ModelPytorch1.2.9 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/399dabb6-49ab-4d74-90d9-7eb8d12c22a9.png" width=128px>gfpganGFP-GAN: Towards Real-World Blind Face Restoration with Generative Facial PriorPytorch1.2.10 及以上JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/1d2dc54c-687d-480e-99eb-130d170a5cd8.png" width=128px>sber-swapSberSwapPytorch1.2.12 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/0a830acf-1653-41c9-9238-79a7a08be07b.png" width=128px>encoder4editingDesigning an Encoder for StyleGAN Image ManipulationPytorch1.2.10 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/f2d70b5c-24b4-4967-9186-4e90ce9e6318.jpg" width=128px>lipganLipGANKeras1.2.15 及以上JP

手势检测

模型参考导出自支持的 Ailia 版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/248f9687-01b1-4ab2-8bb5-86e83123d24a.png" width=128px>yolov3-hand使用 keras-yolov3 的人脸检测分支的手检测Keras1.2.1 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/52ef9890-0abd-4e47-8b3f-0c24f1e963df.jpg" width=128px>hand_detection_pytorchhand-detection.PyTorchPytorch1.2.2 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/21dbe3f2-d3a9-4d4e-90c9-71e1ba074350.png" width=128px>blazepalmMediaPipePyTorchPytorch1.2.5 及以上

手势识别

模型参考导出自支持的 Ailia 版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/461fd726-d5d5-46ad-99d6-7c68c77cfe45.png" width=128px>blazehandMediaPipePyTorchPytorch1.2.5 及以上英文 日文
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/22757409-f246-48ba-ad86-df98e066446e.png" width=128px>hand3dColorHandPose3D networkTensorFlow1.2.5 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/04f3c242-fd69-444e-abbc-0e71d8e1b30e.png" width=128px>minimal-handMinimal HandTensorFlow1.2.8 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/5d20a806-5846-4470-8b8f-252f6c07a8a2.png" width=128px>v2v-posenetV2V-PoseNetPytorch1.2.6 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/10ed8708-5852-4e19-9684-6e3c16a2a811.png" width=128px>hands_segmentation_pytorchhands-segmentation-pytorchPytorch1.2.10 及以上

图像描述

模型参考导出自支持的 Ailia 版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/7d5a398f-33b9-48e3-91d4-2d0119117d0e.jpg" width=128px>illustration2vecIllustration2VecCaffe1.2.2 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/6e558555-c80e-4627-81b7-d83ac4e53bb2.jpg" width=128px>image_captioning_pytorchImage Captioning pytorchPytorch1.2.5 及以上英文 日文
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/08a976f2-24b2-4380-a021-7aadf4822082.png" width=128px>blip2Hugging Face - BLIP-2Pytorch1.2.16 及以上

图像分类

模型参考资料导出自支持的 Ailia 版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/9b37f99e-17b0-4430-9634-09ecbd5405eb.jpg" width=128px>vgg16大型图像识别的非常深度卷积网络Keras1.1.0 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/f0454e5f-c1be-456a-acb3-eab319c72631.jpg" width=128px>googlenet通过卷积深入发展Pytorch1.2.0 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/43061d79-daab-456d-bee6-b30e289b701b.jpg" width=128px>resnet50用于图像识别的深度残差学习Chainer1.2.0 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/c63a0d76-4aa8-49e2-a3ba-230d1a5a1e0e.jpg" width=128px>inceptionv3重新思考计算机视觉的Inception架构Pytorch1.2.0 及以上JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/51babc7e-915d-4983-a8dd-31f44131f028.jpg" width=128px>inceptionv4Keras Inception-V4Keras1.2.5 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/afce2ac2-e09b-4557-a2c0-3bf5dc1c518a.jpg" width=128px>mobilenetv2MobileNet V2的PyTorch实现Pytorch1.2.0 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/37e645c2-1dac-4d14-8807-8ebca5863d2a.jpg" width=128px>mobilenetv3MobileNet V3的PyTorch实现Pytorch1.2.1 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/fae8abe7-5c7c-40f9-81c5-a37e0f2d444f.JPEG" width=128px>partialconv用于填充和图像修复的部分卷积层Pytorch1.2.0 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/48796ae2-cfcc-41bc-b23b-505eab516a83.jpg" width=128px>efficientnetEfficientNet的PyTorch实现Pytorch1.2.3 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/3baf1e49-9c19-4a23-b730-b9826bc920a5.jpg" width=128px>efficientnetv2EfficientNetV2Pytorch1.2.4 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/91718e1a-5d80-4379-822b-ffb1c16f114b.png" width=128px>vitVision Transformer的PyTorch重新实现(图片价值16x16描述:规模识别中的变压器)Pytorch1.2.7 及以上EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/91fc302b-9010-4de9-aeab-0f8cc694a2c9.jpg" width=128px>wide_resnet50宽神经网络Pytorch1.2.5 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/71651497-7c6c-43e6-b164-969f6a32820d.jpg" width=128px>resnet18ResNet18Pytorch1.2.8 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/0b8c18b5-45e4-4875-b9da-8286101cec78.jpg" width=128px>mlp_mixerMLP-MixerPytorch1.2.9 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/153c2d04-57df-438f-8ded-f44c5ee3909c.jpg" width=128px>alexnetAlexNet PyTorchPytorch1.2.5 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/cb6f707b-047e-4536-bf08-40c68a21a054.png" width=128px>clipCLIPPytorch1.2.9 及以上EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/9a17abc3-465d-499d-b9bf-456c40916764.jpeg" width=128px>japanese-clipJapanese-CLIPPytorch1.2.15 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/db4a06dd-69bc-4213-9519-315440ab8c3f.png" width=128px>weather-prediction-from-image从图像预测天气 - (图像的温度)Keras1.2.5 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/f0447cfe-d376-4b00-80e2-993325067597.jpg" width=128px>swin-transformerSwin TransformerPytorch1.2.6 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/76a36ddf-e570-4849-9891-b91fef6ff4a4.jpg" width=128px>convnextConvNeXt的PyTorch实现Pytorch1.2.5 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/35b7900e-9f4b-4a27-9d0e-d0d390c80284.jpg" width=128px>mobileoneMobileOne的PyTorch实现Pytorch1.2.1 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/71500b41-e80d-4d7a-b98b-b7f81294d287.jpg" width=128px>imagenet21kImageNet21KPytorch1.2.11 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/3daa989a-f9bb-448b-a4bb-8e1e1d5de72d.jpeg" width=128px>japanese-stable-clip-vit-l-16japanese-stable-clip-vit-l-16Pytorch1.2.11 及以后版本

图像修复

模型参考导出自支持的Ailia版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/4fdd5cec-25a1-4af5-8c43-647c22a796f3.png" width=128px>部分卷积修复pytorch-inpainting-with-partial-convPyTorch1.2.6及以上EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/e1768d16-3eb4-4f26-94a7-35b036ddf6b1.png" width=128px>Generative多列卷积神经网络修复Image Inpainting via Generative Multi-column Convolutional Neural NetworksTensorFlow1.2.6及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/ed38d764-eeab-45d6-bb92-56ac658695d5.jpg" width=128px>3D照片修复3D Photography using Context-aware Layered Depth InpaintingPytorch1.2.7及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/5a595f6b-4f58-418b-a9c4-12e9d8d11150.png" width=128px>深度填充v2Free-Form Image Inpainting with Gated ConvolutionPytorch1.2.9及以上

图像操作

模型参考导出自支持的Ailia版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/5431b31f-9f69-4632-9250-10bbd469711f.png" width=128px>noise2noiseLearning Image Restoration without Clean DataPytorch1.2.0及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/879128f9-a5f1-4ce3-8c32-b34687cd0495.png" width=128px>dewarpnetDewarpNet: Single-Image Document Unwarping With Stacked 3D and 2D Regression NetworksPytorch1.2.1及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/9c5b80e3-12c1-42c7-8644-05a9aaa7f56c.png" width=128px>illnetDocument Rectification and Illumination Correction using a Patch-based CNNPytorch1.2.2及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/640e1756-d5b3-4ee7-b4a7-2e98c8800b70.jpg" width=128px>colorizationColorful Image ColorizationPytorch1.2.2及以上EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/22ee112f-bf43-486b-93db-0c203838b359.jpg" width=128px>u2net_portraitU^2-Net: Going Deeper with Nested U-Structure for Salient Object DetectionPytorch1.2.2及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/070a2aa0-89c0-4ad6-8bc1-d2cbe226f34d.png" width=128px>style2paintsStyle2PaintsTensorFlow1.2.6及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/c17e4c7c-781d-4ec7-96d7-000245b56b42.png" width=128px>deep_white_balanceDeep White-Balance Editing, CVPR 2020 (Oral)PyTorch1.2.6及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/c2dc19ec-96f4-486e-900e-31e79c4a8393.png" width=128px>deblur_ganDeblurGANPytorch1.2.6及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/0ff05bb6-c1a4-4524-9cde-98b978902272.png" width=128px>可逆去噪网络Invertible Image DenoisingPytorch1.2.8及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/02adf784-f3b0-426b-9f05-228825b9e54d.png" width=128px>dfmDeep Feature MatchingPytorch1.2.6及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/370c20cf-b853-4670-9338-83bb5041b334.png" width=128px>dfeDeep Fundamental Matrix EstimationPytorch1.2.6及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/9592f172-37e5-4d58-b44b-5877422a5f62.png" width=128px>dehamerImage Dehazing Transformer with Transmission-Aware 3D Position EmbeddingPytorch1.2.13及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/21ff5f40-62dd-4414-8b96-a130efc4556f.png" width=128px>pytorch-superpointpytorch-superpoint : Self-Supervised Interest Point Detection and DescriptionPytorch1.2.6及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/d6c8538d-8e38-47a8-bdcf-4c9c65928e0b.png" width=128px>cnngeometric_pytorchCNN几何Pytorch实现Pytorch1.2.7及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/81fc4c84-52b8-4c88-8021-e193d3f3c4cf.png" width=128px>lightglueLightGlue-ONNXPytorch1.2.15及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/d557bd99-4f73-4bdf-83d0-07f91f807003.png" width=128px>docshadowDocShadow-ONNX-TensorRTPytorch1.2.10及以上

图像修复

模型参考导出自支持的Ailia版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/033fc74d-0526-46da-84d2-41964f6c1072.png" width=128px>nafnetNAFNet: Nonlinear Activation Free Network for Image RestorationPytorch1.2.10及以上

图像分割

| | 模型 | 参考 | 导出自 | 支持的 Ailia 版本 | 博客 | |:-----------|------------:|:------------:|:------------:|:------------:|:------------:| | [<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/86482ba2-87bc-4775-b500-084913324d00.png" width=128px>](image_segmentation/deeplabv3/) | [deeplabv3](/image_segmentation/deeplabv3/) | [DeepLab v3+的Xception65主干网络](https://github.com/tensorflow/models/tree/master/research/deeplab) | Chainer | 1.2.0及更高版本 | | [<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/40d6b198-22e0-4a76-b050-ebc2515a97ee.png" width=128px>](image_segmentation/hrnet_segmentation/) | [hrnet_segmentation](/image_segmentation/hrnet_segmentation/) | [高分辨率网络(HRNets)用于语义分割](https://github.com/HRNet/HRNet-Semantic-Segmentation) | Pytorch | 1.2.1及更高版本 | | | [<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/1e320bcc-53eb-406d-aa8a-515fcd22eb59.png" width=128px>](image_segmentation/hair_segmentation/) | [hair_segmentation](/image_segmentation/hair_segmentation/) | [移动设备上的头发分割](https://github.com/thangtran480/hair-segmentation) | Keras | 1.2.1及更高版本 | | | [<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/04966dfb-1ece-4e10-81e3-3889bb017322.png" width=128px>](image_segmentation/pspnet-hair-segmentation/) | [pspnet-hair-segmentation](/image_segmentation/pspnet-hair-segmentation/) | [pytorch-头发分割](https://github.com/YBIGTA/pytorch-hair-segmentation) | Pytorch | 1.2.2及更高版本 | | | [<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/25de63d3-8862-4074-bca6-d44c10cc2650.png" width=128px>](image_segmentation/human_part_segmentation/) | [human_part_segmentation](/image_segmentation/human_part_segmentation/) | [人体解析的自我纠正](https://github.com/PeikeLi/Self-Correction-Human-Parsing) | Pytorch | 1.2.4及更高版本 | [EN](https://medium.com/axinc-ai/humanpartsegmentation-a-machine-learning-model-for-segmenting-human-parts-cd7e39480714) [JP](https://medium.com/axinc/humanpartsegmentation-%E5%8B%95%E7%94%BB%E3%81%8B%E3%82%89%E4%BD%93%E3%81%AE%E9%83%A8%E4%BD%8D%E3%82%92%E3%82%BB%E3%82%B0%E3%83%A1%E3%83%B3%E3%83%86%E3%83%BC%E3%82%B7%E3%83%A7%E3%83%B3%E3%81%99%E3%82%8B%E6%A9%9F%E6%A2%B0%E5%AD%A6%E7%BF%92%E3%83%A2%E3%83%87%E3%83%AB-e8a0e405255) | | [<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/86449775-771e-413d-928f-49cf76ba9ffd.png" width=128px>](image_segmentation/semantic-segmentation-mobilenet-v3/) | [semantic-segmentation-mobilenet-v3](/image_segmentation/semantic-segmentation-mobilenet-v3) | [使用MobileNetV3进行语义分割](https://github.com/OniroAI/Semantic-segmentation-with-MobileNetV3) | TensorFlow | 1.2.5及更高版本 | | | [<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/0e4dd8bc-92a9-4125-bb4c-d30f23e1213b.jpg" width=128px>](image_segmentation/pytorch-unet/) | [pytorch-unet](/image_segmentation/pytorch-unet/) | [Pytorch-Unet](https://github.com/milesial/Pytorch-UNet) | Pytorch | 1.2.5及更高版本 | | | [<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/c181f4bc-2cfa-4804-aeeb-60267951122f.png" width=128px>](image_segmentation/pytorch-enet/) | [pytorch-enet](/image_segmentation/pytorch-enet/) | [PyTorch-ENet](https://github.com/davidtvs/PyTorch-ENet) | Pytorch | 1.2.8及更高版本 | | | [<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/2662f1e7-4ee3-4091-b743-4000ae4335c2.png" width=128px>](image_segmentation/yet-another-anime-segmenter/) | [yet-another-anime-segmenter](/image_segmentation/yet-another-anime-segmenter/) | [Yet Another Anime Segmenter](https://github.com/zymk9/Yet-Another-Anime-Segmenter) | Pytorch | 1.2.6及更高版本 | | | [<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/6dc9b9f1-6ac0-492d-82b7-1317637fd6b6.png" width=128px>](image_segmentation/swiftnet/) | [swiftnet](/image_segmentation/swiftnet/) | [SwiftNet](https://github.com/orsic/swiftnet) | Pytorch | 1.2.6及更高版本 | | | [<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/926da1b5-88a0-4188-bbd4-9043c5afc654.png" width=128px>](image_segmentation/dense_prediction_transformers/) | [dense_prediction_transformers](/image_segmentation/dense_prediction_transformers/) | [用于密集预测的视觉Transformer](https://github.com/intel-isl/DPT) | Pytorch | 1.2.7及更高版本 | [EN](https://medium.com/axinc-ai/dpt-segmentation-model-using-vision-transformer-b479f3027468) [JP](https://medium.com/axinc/dpt-vision-transformer%E3%82%92%E4%BD%BF%E7%94%A8%E3%81%97%E3%81%9F%E3%82%BB%E3%82%B0%E3%83%A1%E3%83%B3%E3%83%86%E3%83%BC%E3%82%B7%E3%83%A7%E3%83%B3%E3%83%A2%E3%83%87%E3%83%AB-88db4842b4a7) | | [<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/5d7ebfb6-aa18-4f59-b800-60637ae534d8.png" width=128px>](image_segmentation/paddleseg/) | [paddleseg](/image_segmentation/paddleseg/) | [PaddleSeg](https://github.com/PaddlePaddle/PaddleSeg/tree/release/2.3/contrib/CityscapesSOTA) | Pytorch | 1.2.7及更高版本 | [EN](https://medium.com/axinc-ai/paddleseg-highly-accurate-segmentation-model-using-hierarchical-attention-18e69363dc2a) [JP](https://medium.com/axinc/paddleseg-%E9%9A%8E%E5%B1%A4%E7%9A%84%E3%81%AA%E3%82%A2%E3%83%86%E3%83%B3%E3%82%B7%E3%83%A7%E3%83%B3%E3%82%92%E4%BD%BF%E7%94%A8%E3%81%97%E3%81%9F%E9%AB%98%E7%B2%BE%E5%BA%A6%E3%81%AA%E3%82%BB%E3%82%B0%E3%83%A1%E3%83%B3%E3%83%86%E3%83%BC%E3%82%B7%E3%83%A7%E3%83%B3%E3%83%A2%E3%83%87%E3%83%AB-acc89bf50423) | | [<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/49c8de46-512a-46e2-bee5-6965a9cb12ee.png" width=128px>](image_segmentation/pp_liteseg/) | [pp_liteseg](/image_segmentation/pp_liteseg/) | [PP-LiteSeg](https://github.com/PaddlePaddle/PaddleSeg/tree/develop/configs/pp_liteseg) | Pytorch | 1.2.10及更高版本 | | | [<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/dabc6212-5b12-46da-8f1b-a0660e55b638.jpg" width=128px>](image_segmentation/suim/) | [suim](/image_segmentation/suim/) | [SUIM](https://github.com/IRVLab/SUIM) | Keras | 1.2.6及更高版本 | | | [<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/724e5404-a272-4eba-afa8-eb7dbf0f3472.png" width=128px>](image_segmentation/group_vit/) | [group_vit](/image_segmentation/group_vit/) | [GroupViT](https://github.com/NVlabs/GroupViT) | Pytorch | 1.2.10及更高版本 | | | [<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/de9b5887-9c5a-4f46-ba5c-519b575b8409.png" width=128px>](image_segmentation/anime-segmentation/) | [anime-segmentation](/image_segmentation/anime-segmentation/) | [动漫分割](https://github.com/SkyTNT/anime-segmentation) | Pytorch | 1.2.9及更高版本 | | [<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/00b27b55-6b7d-45f0-9401-4fa29c4d3b86.png" width=128px>](image_segmentation/segment-anything/) | [segment-anything](/image_segmentation/segment-anything/) | [Segment Anything](https://github.com/facebookresearch/segment-anything) | Pytorch | 1.2.16及更高版本 | | [<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/e3a704dc-e733-40d5-bf3b-047adb2218e3.png" width=128px>](image_segmentation/tusimple-DUC/) | [tusimple-DUC](/image_segmentation/tusimple-DUC/) | [TuSimple-DUC](https://github.com/TuSimple/TuSimple-DUC) | Pytorch | 1.2.10及更高版本 | | | [<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/2e9421ff-ba09-4528-a194-6c68bf5b6ac5.jpg" width=128px>](image_segmentation/pytorch-fcn/) | [pytorch-fcn](/image_segmentation/pytorch-fcn/) | [pytorch-fcn](https://github.com/wkentaro/pytorch-fcn) | Pytorch | 1.3.0及更高版本 | | [<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/fc795dec-3e8f-4a08-b243-75d1b0a7a240.png" width=128px>](image_segmentation/grounded_sam/) | [grounded_sam](/image_segmentation/grounded_sam/) | [Grounded-SAM](https://github.com/IDEA-Research/Grounded-Segment-Anything/tree/main) | Pytorch | 1.2.16及更高版本 | ## 大型语言模型 | 模型 | 参考 | 导出自 | 支持的 Ailia 版本 | 博客 | |------------:|:------------:|:------------:|:------------:|:------------:| |[llava](/large_language_model/llava) | [LLaVA](https://github.com/haotian-liu/LLaVA) | Pytorch | 1.2.16及更高版本 | | ## 地标分类
模型参考文献导出自支持的 Ailia 版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/9518d35b-fdea-43a0-b354-dddf77e2f7fe.jpg" width=128px>landmarks_classifier_asiaLandmarks classifier_asia_V1.1TensorFlow Hub1.2.4 及以上版本EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/b2d10128-1407-45c8-a3c5-dfb7d2372289.jpg" width=128px>places365Release of Places365-CNNsPytorch1.2.5 及以上版本

线段检测

模型参考文献导出自支持的 Ailia 版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/d2e8df5a-37df-4e47-9aa9-6b3e5ecaa796.jpg" width=128px>mlsdM-LSD: Towards Light-weight and Real-time Line Segment DetectionTensorFlow1.2.8 及以上版本EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/91d4b21d-4909-4fb5-ae71-5d6ef8953c97.png" width=128px>dexinedDexiNed: Dense Extreme Inception Network for Edge DetectionPytorch1.2.5 及以上版本

低光照图像增强

模型参考文献导出自支持的 Ailia 版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/f0907d45-21f9-447a-92b8-169020e26f65.png" width=128px>agllnetAGLLNet: Attention Guided Low-light Image Enhancement (IJCV 2021)Pytorch1.2.9 及以上版本EN JP

自然语言处理

| Model | 参考资料 | 导出平台 | 支持的 Ailia 版本 | 博客 | |------------:|:------------:|:------------:|:------------:|:------------:| |[bert](/natural_language_processing/bert) | [pytorch-pretrained-bert](https://pypi.org/project/pytorch-pretrained-bert/) | Pytorch | 1.2.2 或更高 | [EN](https://medium.com/axinc-ai/bert-a-machine-learning-model-for-efficient-natural-language-processing-aef3081c24e8) [JP](https://medium.com/axinc/bert-%E8%87%AA%E7%84%B6%E8%A8%80%E8%AA%9E%E5%87%A6%E7%90%86%E3%82%92%E5%8A%B9%E7%8E%87%E7%9A%84%E3%81%AB%E5%AD%A6%E7%BF%92%E3%81%99%E3%82%8B%E6%A9%9F%E6%A2%B0%E5%AD%A6%E7%BF%92%E3%83%A2%E3%83%87%E3%83%AB-3a9c27d78cf8) | |[bert_maskedlm](/natural_language_processing/bert_maskedlm) | [huggingface/transformers](https://github.com/huggingface/transformers) | Pytorch | 1.2.5 或更高 | |[bert_ner](/natural_language_processing/bert_ner) | [huggingface/transformers](https://github.com/huggingface/transformers) | Pytorch | 1.2.5 或更高 | |[bert_question_answering](/natural_language_processing/bert_question_answering) | [huggingface/transformers](https://github.com/huggingface/transformers) | Pytorch | 1.2.5 或更高 | |[bert_sentiment_analysis](/natural_language_processing/bert_sentiment_analysis) | [huggingface/transformers](https://github.com/huggingface/transformers) | Pytorch | 1.2.5 或更高 | |[bert_zero_shot_classification](/natural_language_processing/bert_zero_shot_classification) | [huggingface/transformers](https://github.com/huggingface/transformers) | Pytorch | 1.2.5 或更高 | |[bert_tweets_sentiment](/natural_language_processing/bert_tweets_sentiment) | [huggingface/transformers](https://github.com/huggingface/transformers) | Pytorch | 1.2.5 或更高 | |[gpt2](/natural_language_processing/gpt2) | [GPT-2](https://github.com/onnx/models/blob/master/text/machine_comprehension/gpt-2/README.md) | Pytorch | 1.2.7 或更高 | |[rinna_gpt2](/natural_language_processing/rinna_gpt2) | [japanese-pretrained-models](https://github.com/rinnakk/japanese-pretrained-models) | Pytorch | 1.2.7 或更高 | |[fugumt-en-ja](/natural_language_processing/fugumt-en-ja) | [Fugu-Machine Translator](https://github.com/s-taka/fugumt) | Pytorch | 1.2.9 或更高 | [JP](https://medium.com/axinc/fugumt-%E8%8B%B1%E8%AA%9E%E3%81%8B%E3%82%89%E6%97%A5%E6%9C%AC%E8%AA%9E%E3%81%B8%E3%81%AE%E7%BF%BB%E8%A8%B3%E3%82%92%E8%A1%8C%E3%81%86%E6%A9%9F%E6%A2%B0%E5%AD%A6%E7%BF%92%E3%83%A2%E3%83%87%E3%83%AB-46b839c1b4ae) | |[fugumt-ja-en](/natural_language_processing/fugumt-ja-en) | [Fugu-Machine Translator](https://github.com/s-taka/fugumt) | Pytorch | 1.2.10 或更高 | |[bert_sum_ext](/natural_language_processing/bert_sum_ext) | [BERTSUMEXT](https://github.com/dmmiller612/bert-extractive-summarizer) | Pytorch | 1.2.7 或更高 | |[sentence_transformers_japanese](/natural_language_processing/sentence_transformers_japanese) | [sentence transformers](https://huggingface.co/sentence-transformers/paraphrase-multilingual-mpnet-base-v2) | Pytorch | 1.2.7 或更高 | [JP](https://medium.com/axinc/sentencetransformer-%E3%83%86%E3%82%AD%E3%82%B9%E3%83%88%E3%81%8B%E3%82%89embedding%E3%82%92%E5%8F%96%E5%BE%97%E3%81%99%E3%82%8B%E8%A8%80%E8%AA%9E%E5%87%A6%E7%90%86%E3%83%A2%E3%83%87%E3%83%AB-b7d2a9bb2c31) | |[presumm](/natural_language_processing/presumm) | [PreSumm](https://github.com/nlpyang/PreSumm) | Pytorch | 1.2.8 或更高| |[t5_base_japanese_title_generation](/natural_language_processing/t5_base_japanese_title_generation) | [t5-japanese](https://github.com/sonoisa/t5-japanese) | Pytorch | 1.2.13 或更高 | [JP](https://medium.com/axinc/t5-%E3%83%86%E3%82%AD%E3%82%B9%E3%83%88%E3%81%8B%E3%82%89%E3%83%86%E3%82%AD%E3%82%B9%E3%83%88%E3%82%92%E7%94%9F%E6%88%90%E3%81%99%E3%82%8B%E6%A9%9F%E6%A2%B0%E5%AD%A6%E7%BF%92%E3%83%A2%E3%83%87%E3%83%AB-602830bdc5b4) | |[bertjsc](/natural_language_processing/bertjsc) | [bertjsc](https://github.com/er-ri/bertjsc) | Pytorch | 1.2.15 或更高 | |[multilingual-e5](/natural_language_processing/multilingual-e5) | [multilingual-e5-base](https://huggingface.co/intfloat/multilingual-e5-base) | Pytorch | 1.2.15 或更高 | [JP](https://medium.com/axinc/multilingual-e5-%E5%A4%9A%E8%A8%80%E8%AA%9E%E3%81%AE%E3%83%86%E3%82%AD%E3%82%B9%E3%83%88%E3%82%92embedding%E3%81%99%E3%82%8B%E6%A9%9F%E6%A2%B0%E5%AD%A6%E7%BF%92%E3%83%A2%E3%83%87%E3%83%AB-71f1dec7c4f0) | |[bert_insert_punctuation](/natural_language_processing/bert_insert_punctuation) | [bert-japanese](https://github.com/cl-tohoku/bert-japanese) | Pytorch | 1.2.15 或更高 | |[t5_whisper_medical](/natural_language_processing/t5_whisper_medical) | 使用 t5 进行医学术语纠错 | Pytorch | 1.2.13 或更高 | | |[t5_base_summarization](/natural_language_processing/t5_base_japanese_summarization) | [t5-japanese](https://github.com/sonoisa/t5-japanese) | Pytorch | 1.2.13 或更高 | |[glucose](/natural_language_processing/glucose) | [GLuCoSE (General Luke-based Contrastive Sentence Embedding)-base-Japanese](https://huggingface.co/pkshatech/GLuCoSE-base-ja) | Pytorch | 1.2.15 或更高 | |[cross_encoder_mmarco](/natural_language_processing/cross_encoder_mmarco) | [jeffwan/mmarco-mMiniLMv2-L12-H384-v](https://huggingface.co/jeffwan/mmarco-mMiniLMv2-L12-H384-v1) | Pytorch | 1.2.10 或更高 | [JP](https://medium.com/axinc/crossencodermmarco-%E8%B3%AA%E5%95%8F%E6%96%87%E3%81%A8%E5%9B%9E%E7%AD%94%E6%96%87%E3%81%AE%E9%A1%9E%E4%BC%BC%E5%BA%A6%E3%82%92%E8%A8%88%E7%AE%97%E3%81%99%E3%82%8B%E6%A9%9F%E6%A2%B0%E5%AD%A6%E7%BF%92%E3%83%A2%E3%83%87%E3%83%AB-c90b35e9fc09)| |[soundchoice-g2p](/natural_language_processing/soundchoice-g2p) | [Hugging Face - speechbrain/soundchoice-g2p](https://huggingface.co/speechbrain/soundchoice-g2p) | Pytorch | 1.2.16 或更高 | | |[g2p_en](/natural_language_processing/g2p_en) | [g2p_en](https://github.com/Kyubyong/g2p) | Pytorch | 1.2.14 或更高 | | |[t5_base_japanese_ner](/natural_language_processing/t5_base_japanese_ner) | [t5-japanese](https://github.com/sonoisa/t5-japanese) | Pytorch | 1.2.13 或更高 | |[japanese-reranker-cross-encoder](/natural_language_processing/japanese-reranker-cross-encoder) | [hotchpotch/japanese-reranker-cross-encoder-large-v1](https://huggingface.co/hotchpotch/japanese-reranker-cross-encoder-large-v1) | Pytorch | 1.2.16 或更高 | ## 网络入侵检测 | Model | 参考资料 | 导出平台 | 支持的 Ailia 版本 | 博客 | |------------:|:------------:|:------------:|:------------:|:------------:| | [bert-network-packet-flow-header-payload](/network_intrusion_detection/bert-network-packet-flow-header-payload/) | [bert-network-packet-flow-header-payload](https://huggingface.co/rdpahalavan/bert-network-packet-flow-header-payload)| Pytorch | 1.2.10 或更高 | | | [falcon-adapter-network-packet](/network_intrusion_detection/falcon-adapter-network-packet/) | [falcon-adapter-network-packet](https://huggingface.co/rdpahalavan/falcon-adapter-network-packet)| Pytorch | 1.2.10 或更高 | | ## 神经渲染 | | Model | 参考资料 | 导出平台 | 支持的 Ailia 版本 | 博客 | |:-----------|------------:|:------------:|:------------:|:------------:|:------------:| | [<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/0f455f74-1f46-4bb6-a0b4-d2abb0164621.png" width=128px>](neural_rendering/nerf/) | [nerf](/neural_rendering/nerf/) | [NeRF: Neural Radiance Fields](https://github.com/bmild/nerf) | Tensorflow | 1.2.10 或更高 | [EN](https://medium.com/axinc-ai/nerf-machine-learning-model-to-generate-and-render-3d-models-from-multiple-viewpoint-images-599631dc2075) [JP](https://medium.com/axinc/nerf-%E8%A4%87%E6%95%B0%E3%81%AE%E8%A6%96%E7%82%B9%E3%81%AE%E7%94%BB%E5%83%8F%E3%81%8B%E3%82%893d%E3%83%A2%E3%83%87%E3%83%AB%E3%82%92%E7%94%9F%E6%88%90%E3%81%97%E3%81%A6%E3%83%AC%E3%83%B3%E3%83%80%E3%83%AA%E3%83%B3%E3%82%B0%E3%81%99%E3%82%8B%E6%A9%9F%E6%A2%B0%E5%AD%A6%E7%BF%92%E3%83%A2%E3%83%87%E3%83%AB-2d6bee7ff22f) | | [<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/e34e7237-6308-4e49-b2a9-ba04fbf94180.gif" width=128px>](neural_rendering/tripo_sr/) | [TripoSR](/neural_rendering/tripo_sr/) | [TripoSR](https://github.com/VAST-AI-Research/TripoSR) | Pytorch | 1.2.6 或更高 | ## 不适宜内容检测
模型参考资料导出自支持的 Ailia 版本博客
clip-based-nsfw-detectorCLIP-based-NSFW-DetectorKeras1.2.10 及之后版本JP

物体检测

模型参考来源支持的 Ailia 版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/0526e23d-54c7-4966-ba0a-53c148994d0e.png" width=128px>yolov1-tinyYOLO: 实时物体检测Darknet1.1.0 及之后JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/bb35484b-fc4f-466b-b89e-b4c727d065b1.png" width=128px>yolov2YOLO: 实时物体检测Pytorch1.2.0 及之后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/456eae46-4a60-44f8-80f4-d9bf84274b11.png" width=128px>yolov2-tinyYOLO: 实时物体检测Pytorch1.2.6 及之后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/9398e953-3be2-4f7c-87b7-9576e5065557.png" width=128px>yolov3YOLO: 实时物体检测ONNX Runtime1.2.1 及之后EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/051b1164-fa58-4b9c-90a5-28b46c04db8f.png" width=128px>yolov3-tinyYOLO: 实时物体检测ONNX Runtime1.2.1 及之后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/125d34ea-5ddb-4ef5-b4ef-c8790d3740f8.png" width=128px>yolov4Pytorch-YOLOv4Pytorch1.2.4 及之后EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/a9af46e5-0fac-4f2d-8bfc-ff2eea91d905.png" width=128px>yolov4-tinyPytorch-YOLOv4Pytorch1.2.5 及之后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/384b459f-22df-4a85-8296-17325f952ff5.png" width=128px>yolov5yolov5Pytorch1.2.5 及之后EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/ac28b494-a2e3-4203-931e-079e9264fd61.jpg" width=128px>yolov6YOLOV6Pytorch1.2.10 及之后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/367f57da-6670-4c24-99c2-748b2e38e87a.jpg" width=128px>yolov7YOLOv7Pytorch1.2.7 及之后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/5413b4b5-3d46-4b05-b910-f8fd3982b7e8.png" width=128px>yolov8YOLOv8Pytorch1.2.14.1 及之后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/bdbb63e2-0581-4a2a-9da2-c61a5f6ad9c2.png" width=128px>yolov8-segYOLOv8Pytorch1.2.14.1 及之后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/c25c01f6-0a1b-4a0b-b96b-f6e7c1677b99.png" width=128px>yolov9YOLOv9Pytorch1.2.10 及之后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/876aa939-8c24-4e95-95c4-eca255fc2c97.jpg" width=128px>yoloryolorPytorch1.2.5 及之后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/a549c213-179b-4157-b3f7-23dd04e06f4d.jpg" width=128px>yoloxYOLOXPytorch1.2.6 及之后EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/6b5b1d63-3b73-486e-91fb-80901f977c43.jpg" width=128px>yolox-ti-liteedgeai-yoloxPytorch1.2.9 及之后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/890259c3-d37b-418b-9e05-9959c7c04363.jpg" width=128px>yolovYOLOVPytorch1.2.10 及之后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/4fde6651-85c7-4310-a26b-24792a642652.png" width=128px>mobilenet_ssdMobileNetV1, MobileNetV2, VGG based SSD/SSD-lite implementation in PytorchPytorch1.2.1 及之后EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/38bb4914-bb32-4824-ad80-f338e023286a.png" width=128px>maskrcnnMask R-CNN: 实时神经网络物体实例分割Pytorch1.2.3 及之后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/c39b140f-5b27-489c-ae5a-d83f74f033b6.png" width=128px>m2detM2Det: 基于多层次特征金字塔网络的单阶段物体检测器Pytorch1.2.3及更高版本EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/ad03e8b1-3a9b-4575-aac2-8817ee523ee8.png" width=128px>centernetCenterNet: 将物体视为点Pytorch1.2.1及更高版本EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/df59ba2d-f223-49ea-8213-4d9e277a175d.png" width=128px>pedestrian_detectionPedestrian-Detection-on-YOLOv3_Research-and-APPKeras1.2.1及更高版本
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/d4be1c66-0b08-4c48-9fbe-9d3530be865b.png" width=128px>efficientdetEfficientDet: 在PyTorch中可扩展且高效的物体检测Pytorch1.2.6及更高版本
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/db13a97f-b633-4f93-8557-7816f6528d50.jpg" width=128px>nanodetNanoDetPytorch1.2.6及更高版本
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/f13ed5d1-e4ba-4edf-ad33-de0132933fd7.png" width=128px>mobile_object_localizermobile_object_localizer_v1TensorFlow Hub1.2.6及更高版本EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/adc70fc6-c742-4153-9b85-68848e51af68.png" width=128px>sku110k-densedetSKU110K-DenseDetPytorch1.2.9及更高版本EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/2f110bf7-b1a1-4deb-a861-896104d7806e.png" width=128px>traffic-sign-detection交通标志检测Tensorflow1.2.10及更高版本EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/3c1bbe82-25e1-49f9-a364-ccbefe84cd43.png" width=128px>detic使用图像级监督检测两万个类别Pytorch1.2.10及更高版本EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/3be59133-9cc3-4773-ab53-5e2473aef1de.png" width=128px>picodetPP-PicoDetPytorch1.2.10及更高版本
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/3ccb844a-dac7-4a48-bc30-eb6391e8cce6.png" width=128px>yolactYou Only Look At CoefficienTsPytorch1.2.6及更高版本
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/b7526841-c3f4-4bf3-9952-5467b830f711.png" width=128px>fastest-detFastestDetPytorch1.2.5及更高版本
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/68143bb1-7b04-4322-8dc5-6e609abbf6a4.jpg" width=128px>dab-detrDAB-DETRPytorch1.2.12及更高版本
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/691c1818-5983-430b-a650-ff605a323778.jpg" width=128px>poly_yoloPoly YOLOKeras1.2.6及更高版本
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/c299f028-84ca-4354-9f66-be1c675d7df2.png" width=128px>glipGLIPPytorch1.2.13及更高版本
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/e9d4dfe3-20fc-49d6-90be-5686249a057a.png" width=128px>crowd_det人群场景中的检测Pytorch1.2.13及更高版本
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/c5483e1e-b9f3-43ff-b73c-6bf8960f9637.png" width=128px>footandballFootAndBall: 集成的球员和球检测器Pytorch1.2.0及更高版本
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/f0090f54-2ffe-4fc5-ae29-ed9f29a43fff.jpg" width=128px>qrcode_wechatqrcodeqrcode_wechatqrcodeCaffe1.2.15及更高版本
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/0b73e88b-1227-410c-806d-5ba0b06bc191.jpg" width=128px>layout_parsingunstructured-inferencePytorch1.2.9及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/244d0fe9-032b-494b-8625-748d4a69aacb.jpg" width=128px>damo_yoloDAMO-YOLOPytorch1.2.9及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/4b4adfbf-10be-485c-98d0-546a09fa23e9.png" width=128px>groundingdinoGrounding DINOPytorch1.2.16及以上

目标检测 3D

模型参考导出自支持的 Ailia 版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/d494f3ce-bf21-4033-bec1-929e48ae781f.png" width=128px>3d_bbox使用深度学习和几何学进行3D边界框估计Pytorch1.2.6 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/5814920f-16c3-4787-ab89-581f1beae2c7.png" width=128px>3d-object-detection.pytorch3d-object-detection.pytorchPytorch1.2.8 及以上EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/072a8be0-6753-4b29-a2eb-402ec35f3337.png" width=128px>mediapipe_objectronMediaPipe ObjectronTensorFlow Lite1.2.5 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/19e79ed4-06b3-4599-974c-ac2989655336.png" width=128px>egonetEgoNetPytorch1.2.9 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/9657366b-d311-4e35-9885-a4d6d91ebecd.png" width=128px>d4lcnD4LCNPytorch1.2.9 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/0248c52b-9f61-4c72-853d-e2adbeee8bd8.png" width=128px>did_m3dDID M3DPytorch1.2.11 及以上

目标跟踪

模型参考导出自支持的 Ailia 版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/d614fe8b-b0dd-46ac-9865-909125faf425.gif" width=128px>deepsort使用 PyTorch 的 Deep SortPytorch1.2.3 及以上EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/72acc7ca-a1cd-4240-8b00-326c082013e0.png" width=128px>person_reid_baseline_pytorchUTS-Person-reID-PracticalPytorch1.2.6 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/a656be6b-6473-41b1-ae88-86c422aa44ec.png" width=128px>abd_net注意但多样的人员再识别Pytorch1.2.7 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/53f74070-e589-444f-9271-3e8b8f45c13c.png" width=128px>siam-motSiamMOTPytorch1.2.9 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/6aa83ddf-9781-43f2-8ca8-91e5e8bfcf31.png" width=128px>bytetrackByteTrackPytorch1.2.5 及以上EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/53901ac1-54f0-4679-9b15-9fcdedd79d20.png" width=128px>qd-3dt单目准密3D目标跟踪Pytorch1.2.11 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/dc2f8055-b959-43d8-9f92-d15dd3310eb5.png" width=128px>strong_sortStrongSORTPytorch1.2.15 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/ab75bca6-3c5d-4a02-9bbb-4edc600ba0f0.png" width=128px>centroids-reid质心在图像检索中的非同寻常效果Pytorch1.2.9 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/bc4ab7d2-73c6-404a-908b-7abdf26a82bf.jpg" width=128px>deepsort_vehicle多摄像头实时目标跟踪Pytorch1.2.9 及以上

光流估计

模型参考导出自支持的 Ailia 版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/5d3d31e4-37ce-487a-8baa-a1475e0b5a2a.png" width=128px>raftRAFT: 光流的循环所有对字段变换Pytorch1.2.6 及以上EN JP

点云分割

模型参考导出自支持的 Ailia 版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/f737e857-1a03-4211-9404-0aaefe153ac4.png" width=128px>pointnet_pytorchPointNet.pytorchPytorch1.2.6 及以上

姿态估计

模型参考导出自支持的Ailia版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/9f0587f8-22f1-483b-bf82-a4242cdf1936.png" width=128px>openposeCVPR'17(口头)实时多人姿态估计代码库Caffe1.2.1及以后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/a4f99989-ee3d-4ace-8eaa-51772dda207a.png" width=128px>lightweight-human-pose-estimationPyTorch中快速准确的姿态估计。<br/>包含“轻量级OpenPose的实时CPU多人2D姿态估计”论文的实现。Pytorch1.2.1及以后EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/81e81633-78fd-459b-98ba-127758629c80.png" width=128px>pose_resnet人类姿态估计和跟踪的简单基线Pytorch1.2.1及以后EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/b68a6889-4bd5-42ed-9360-9a61af1838af.png" width=128px>blazeposeMediaPipePyTorchPytorch1.2.5及以后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/e9d650c3-7951-4537-997b-53d88edac2fb.png" width=128px>efficientposeEfficientPose代码库TensorFlow1.2.6及以后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/c6870c6e-2fa4-4d4b-9764-4061c0a4bd7b.png" width=128px>movenetmovenet代码库TensorFlow1.2.8及以后EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/14081db4-0359-4009-8cf9-1403888e1395.png" width=128px>animalposeMMPose - 2D动物姿态估计Pytorch1.2.7及以后EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/d407740c-4c52-408e-8f37-0f1a5c36c184.png" width=128px>mediapipe_holisticMediaPipe HolisticTensorFlow1.2.9及以后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/1a887d6b-614d-48d6-8605-2c1c73d1e687.png" width=128px>ap-10kAP-10KPytorch1.2.4及以后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/2427d760-cbf6-4187-a04a-a3e5152244d1.png" width=128px>posenetPoseNet PytorchPytorch1.2.10及以后
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/159c8577-d71b-4c3a-b9e7-94bafda62230.png" width=128px>e2poseE2PoseTensorflow1.2.5及以后

3D姿态估计

            | | 模型 | 参考 | 导出自 | 支持的 Ailia 版本 | 博客 |

|:-----------|------------:|:------------:|:------------:|:------------:|:------------:| | <img src="https://yellow-cdn.veclightyear.com/35dd4d3f/64ff684f-a9b3-46a8-8256-8d6ff791fb4d.png" width=128px> |lightweight-human-pose-estimation-3d | PyTorch 实时 3D 多人姿态估计 demo。<br/>OpenVINO 后端可用于快速的 CPU 推理。 | Pytorch | 1.2.1及更高 | | | <img src="https://yellow-cdn.veclightyear.com/35dd4d3f/920eeeb7-f450-4553-98d9-5ff039a7122d.png" width=128px> |3d-pose-baseline | 一个简单的 TensorFlow 3D 人体姿态估计基线。<br/>在 ICCV 17 上展示。 | TensorFlow | 1.2.3及更高 | | | <img src="https://yellow-cdn.veclightyear.com/35dd4d3f/361f597b-cfa5-4f40-add2-345965eba261.png" width=128px> |pose-hg-3d | Towards 3D Human Pose Estimation in the Wild: a Weakly-supervised Approach | Pytorch | 1.2.6及更高 | | | <img src="https://yellow-cdn.veclightyear.com/35dd4d3f/c6db3dae-f60e-4ffd-a00b-f54d7526af34.png" width=128px> |blazepose-fullbody | MediaPipe | TensorFlow Lite | 1.2.5及更高 | EN JP | | <img src="https://yellow-cdn.veclightyear.com/35dd4d3f/7b9fef90-63d1-4529-af2f-2ad4d254eaa3.jpg" width=128px> |3dmppe_posenet | 来自单个 RGB 图像的 3D 多人姿态估计的“相机距离感知自上而下方法”的 PoseNet | Pytorch | 1.2.6及更高 | | | <img src="https://yellow-cdn.veclightyear.com/35dd4d3f/df886dbe-9b3a-484e-9ee6-138bc992f66e.png" width=128px> |gast | 视频中 3D 人体姿态估计的图注意空间时间卷积网络 (GAST-Net) | Pytorch | 1.2.7及更高 | EN JP | | <img src="https://yellow-cdn.veclightyear.com/35dd4d3f/1fd4dc80-46ad-44a8-88bb-06d5a2958996.png" width=128px> |mediapipe_pose_world_landmarks | MediaPipe 实际坐标点 | TensorFlow Lite | 1.2.10及更高 | |

道路检测

模型参考导出自支持的 Ailia 版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/0bb6af51-8160-46f6-b301-4773aa903307.jpg" width=128px>codes-for-lane-detection车道检测代码Pytorch1.2.6及更高EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/a77fe165-925c-44a0-85ce-ee079187a5ab.jpg" width=128px>roneldRONELD 车道检测Pytorch1.2.6及更高
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/381b35e1-2c40-4d5e-aa1a-4f74a965fc90.png" width=128px>road-segmentation-adasroad-segmentation-adas-0001OpenVINO1.2.5及更高
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/82fa90b4-884d-4499-aa21-21ea03d38ca2.png" width=128px>cdnetCDNetPytorch1.2.5及更高
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/7995e10b-667d-4eae-9f55-efd3e45971eb.jpg" width=128px>lstrLSTRPytorch1.2.8及更高
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/f91e749c-8a37-47c3-9881-ee3a99200ee1.jpg" width=128px>ultra-fast-lane-detectionUltra-Fast-Lane-DetectionPytorch1.2.6及更高
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/664f03bd-36e1-479b-82d0-44e16120241d.jpg" width=128px>yolopYOLOPPytorch1.2.6及更高
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/43ea3e50-eac4-4a2e-92b5-9f01d3471fde.jpg" width=128px>hybridnetsHybridNetsPytorch1.2.6及更高
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/26bc147c-8433-4458-a829-c03c5034736b.jpg" width=128px>polylanenetPolyLaneNetPytorch1.2.9及更高

旋转预测

模型参考导出自支持的 Ailia 版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/b17f194b-7696-4b17-a47d-d52eb33f0eba.png" width=256px>rotnet用于预测图像旋转角度以校正其方向的 CNNKeras1.2.1及更高

样式迁移

模型参考文献来源支持的 Ailia 版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/263ebe21-1c66-4cd0-87c7-db5ec6f1c3c0.png" width=128px>adain实时任意风格转换与自适应实例归一化Pytorch1.2.1 及以上EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/01190eda-7d73-4f7a-97d9-ac46d37fc610.png" width=128px>psganPSGAN: 姿势和表情鲁棒的空间感知 GAN,用于定制化妆转换Pytorch1.2.7 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/cd1b962b-0a66-4c71-9f9c-e18960f093bb.png" width=128px>beauty_ganBeautyGANPytorch1.2.7 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/446c50a6-7586-4781-9045-e8c10347a90c.png" width=128px>animeganv2AnimeGANv2 的 PyTorch 实现Pytorch1.2.5 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/d06cea6e-0401-4577-9727-e847f8e2f38e.png" width=128px>pix2pixHDpix2pixHD: 使用条件 GAN 进行高分辨率图像合成和语义操控Pytorch1.2.6 及以上

超分辨率

模型参考文献来源支持的 Ailia 版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/68526584-4fab-49b2-82e1-7d8a253c40e0.png" width=128px>srresnet用生成对抗网络实现照片级单图像超分辨率Pytorch1.2.0 及以上EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/b871bfef-9f89-4ce4-be52-d6d0295c1099.png" width=128px>edsr用于单图像超分辨率的增强深度残差网络Pytorch1.2.6 及以上EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/d7361ee6-2088-4adc-9b5e-bbf5ac147ba3.png" width=128px>han通过整体注意网络实现单图像超分辨率Pytorch1.2.6 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/1289c344-c962-4d75-8695-5b8bc3418568.jpg" width=128px>real-esrganReal-ESRGANPytorch1.2.9 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/7388c3b7-9531-4552-af93-df72d3891750.png" width=128px>rcan-it重新审视 RCAN: 改进的单图像超分辨率训练Pytorch1.2.10 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/0ba3d0c8-347b-40b7-a91b-d66dc001138a.png" width=128px>swinirSwinIR: 使用 Swin Transformer 的图像恢复Pytorch1.2.12 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/d0d78b03-9a41-4c37-bdb5-546b26f6e447.png" width=128px>HatHatPytorch1.2.6 及以上

文本检测

模型参考文献来源支持的 Ailia 版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/35a9ae18-157e-4348-a498-bf3508c539de.jpg" width=64px>craft_pytorchCRAFT: 文本检测的字符区域感知Pytorch1.2.2 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/115eadfd-7f95-480b-8525-2a6d0a3b9607.png" width=64px>pixel_linkPixel-LinkTensorFlow1.2.6 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/deb32505-ac19-41d1-84b6-4bc155b48283.png" width=64px>eastEAST: 一种高效且准确的场景文本检测器TensorFlow1.2.6 及以上

文本识别

模型参考导出自支持的 Ailia 版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/38e44fa3-060c-458f-b40b-b500b8c164da.png" width=64px>etl日本字符分类Keras1.1.0 及以上JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/b5635e60-e8c5-4f92-bcac-e50a67575118.png" width=64px>deep-text-recognition-benchmarkdeep-text-recognition-benchmarkPytorch1.2.6 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/c5e8eb2a-6941-4574-8060-9d619af3d035.png" width=64px>crnn.pytorch卷积递归神经网络Pytorch1.2.6 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/7ed1256e-de4e-4371-89b6-aa004b7b878d.png" width=64px>paddleocrPaddleOCR:基于 PaddlePaddle 的多语言 OCR 工具包Pytorch1.2.6 及以上EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/e8f5245e-b426-4ae9-afa7-2f49c34367ec.jpg" width=64px>easyocr支持 80 多种语言的即用型 OCRPytorch1.2.6 及以上
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/ae060149-ab03-44e1-98b8-f041f66fff9a.png" width=64px>ndlocr_text_recognitionNDL OCRPytorch1.2.5 及以上

时间序列预测

模型参考导出自支持的 Ailia 版本博客
informer2020Informer:用于长序列时间序列预测的高效转换器 (AAAI'21 最佳论文)Pytorch1.2.10 及以上

车辆识别

模型参考导出自支持的 Ailia 版本博客
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/c3750943-c807-4b51-a226-18a698d21238.png" width=64px>vehicle-attributes-recognition-barriervehicle-attributes-recognition-barrier-0042OpenVINO1.2.5 及以上EN JP
<img src="https://yellow-cdn.veclightyear.com/35dd4d3f/e7ad7fae-93b9-4108-b2ae-3aec4099cd93.png" width=128px>vehicle-license-plate-detection-barriervehicle-license-plate-detection-barrier-0106OpenVINO1.2.5 及以上

商业模型

模型参考导出自支持的 Ailia 版本博客
acculus-poseAcculus, Inc.Caffe1.2.3 及以上

其他语言

编辑推荐精选

讯飞智文

讯飞智文

一键生成PPT和Word,让学习生活更轻松

讯飞智文是一个利用 AI 技术的项目,能够帮助用户生成 PPT 以及各类文档。无论是商业领域的市场分析报告、年度目标制定,还是学生群体的职业生涯规划、实习避坑指南,亦或是活动策划、旅游攻略等内容,它都能提供支持,帮助用户精准表达,轻松呈现各种信息。

AI办公办公工具AI工具讯飞智文AI在线生成PPTAI撰写助手多语种文档生成AI自动配图热门
讯飞星火

讯飞星火

深度推理能力全新升级,全面对标OpenAI o1

科大讯飞的星火大模型,支持语言理解、知识问答和文本创作等多功能,适用于多种文件和业务场景,提升办公和日常生活的效率。讯飞星火是一个提供丰富智能服务的平台,涵盖科技资讯、图像创作、写作辅助、编程解答、科研文献解读等功能,能为不同需求的用户提供便捷高效的帮助,助力用户轻松获取信息、解决问题,满足多样化使用场景。

热门AI开发模型训练AI工具讯飞星火大模型智能问答内容创作多语种支持智慧生活
Spark-TTS

Spark-TTS

一种基于大语言模型的高效单流解耦语音令牌文本到语音合成模型

Spark-TTS 是一个基于 PyTorch 的开源文本到语音合成项目,由多个知名机构联合参与。该项目提供了高效的 LLM(大语言模型)驱动的语音合成方案,支持语音克隆和语音创建功能,可通过命令行界面(CLI)和 Web UI 两种方式使用。用户可以根据需求调整语音的性别、音高、速度等参数,生成高质量的语音。该项目适用于多种场景,如有声读物制作、智能语音助手开发等。

Trae

Trae

字节跳动发布的AI编程神器IDE

Trae是一种自适应的集成开发环境(IDE),通过自动化和多元协作改变开发流程。利用Trae,团队能够更快速、精确地编写和部署代码,从而提高编程效率和项目交付速度。Trae具备上下文感知和代码自动完成功能,是提升开发效率的理想工具。

AI工具TraeAI IDE协作生产力转型热门
咔片PPT

咔片PPT

AI助力,做PPT更简单!

咔片是一款轻量化在线演示设计工具,借助 AI 技术,实现从内容生成到智能设计的一站式 PPT 制作服务。支持多种文档格式导入生成 PPT,提供海量模板、智能美化、素材替换等功能,适用于销售、教师、学生等各类人群,能高效制作出高品质 PPT,满足不同场景演示需求。

讯飞绘文

讯飞绘文

选题、配图、成文,一站式创作,让内容运营更高效

讯飞绘文,一个AI集成平台,支持写作、选题、配图、排版和发布。高效生成适用于各类媒体的定制内容,加速品牌传播,提升内容营销效果。

热门AI辅助写作AI工具讯飞绘文内容运营AI创作个性化文章多平台分发AI助手
材料星

材料星

专业的AI公文写作平台,公文写作神器

AI 材料星,专业的 AI 公文写作辅助平台,为体制内工作人员提供高效的公文写作解决方案。拥有海量公文文库、9 大核心 AI 功能,支持 30 + 文稿类型生成,助力快速完成领导讲话、工作总结、述职报告等材料,提升办公效率,是体制打工人的得力写作神器。

openai-agents-python

openai-agents-python

OpenAI Agents SDK,助力开发者便捷使用 OpenAI 相关功能。

openai-agents-python 是 OpenAI 推出的一款强大 Python SDK,它为开发者提供了与 OpenAI 模型交互的高效工具,支持工具调用、结果处理、追踪等功能,涵盖多种应用场景,如研究助手、财务研究等,能显著提升开发效率,让开发者更轻松地利用 OpenAI 的技术优势。

Hunyuan3D-2

Hunyuan3D-2

高分辨率纹理 3D 资产生成

Hunyuan3D-2 是腾讯开发的用于 3D 资产生成的强大工具,支持从文本描述、单张图片或多视角图片生成 3D 模型,具备快速形状生成能力,可生成带纹理的高质量 3D 模型,适用于多个领域,为 3D 创作提供了高效解决方案。

3FS

3FS

一个具备存储、管理和客户端操作等多种功能的分布式文件系统相关项目。

3FS 是一个功能强大的分布式文件系统项目,涵盖了存储引擎、元数据管理、客户端工具等多个模块。它支持多种文件操作,如创建文件和目录、设置布局等,同时具备高效的事件循环、节点选择和协程池管理等特性。适用于需要大规模数据存储和管理的场景,能够提高系统的性能和可靠性,是分布式存储领域的优质解决方案。

下拉加载更多