paper-with-me

Action Classification 벤치마크

Action Classification on MiT

29개 결과 · ⬇ CSV · JSON

Top 1 Accuracy

27.5 33.9 40.3 46.7 53.1 2017-05 2026-09 I3D — 29.51 (2017-05-22) TRN-Multiscale — 28.27 (2017-11-22) EvaNet — 31.8 (2018-11-26) AssembleNet — 34.27 (2019-05-30) CoST (ResNet-101, 32 frames) — 32.4 (2019-06-01) SRTG r3d-101 — 33.56 (2020-06-15) SRTG r(2+1)d-50 — 31.6 (2020-06-15) SRTG r3d-50 — 30.72 (2020-06-15) SRTG r(2+1)d-34 — 28.97 (2020-06-15) SRTG r3d-34 — 28.55 (2020-06-15) VTN — 37.4 (2021-02-01) MoViNet-A6 — 40.2 (2021-03-21) MoViNet-A5 — 39.1 (2021-03-21) MoViNet-A4 — 37.9 (2021-03-21) MoViNet-A3 — 35.6 (2021-03-21) MoViNet-A2 — 34.3 (2021-03-21) MoViNet-A1 — 32.0 (2021-03-21) MoViNet-A0 — 27.5 (2021-03-21) VATT-Large — 41.1 (2021-04-22) MBT (AV) — 37.3 (2021-06-30) CoVeR(JFT-3B) — 46.1 (2021-12-14) CoVeR(JFT-300M) — 45.0 (2021-12-14) MTV-H (WTS 60M) — 47.2 (2022-01-12) UniFormerV2-L — 47.8 (2022-09-22) UMT-L (ViT-L/16) — 48.7 (2023-03-28) OmniVec2 — 53.1 (2024-01-01) InternVideo2-1B — 50.9 (2024-03-22) I3D — 29.51 (2017-05-22) EvaNet — 31.8 (2018-11-26) AssembleNet — 34.27 (2019-05-30) VTN — 37.4 (2021-02-01) MoViNet-A6 — 40.2 (2021-03-21) VATT-Large — 41.1 (2021-04-22) CoVeR(JFT-3B) — 46.1 (2021-12-14) MTV-H (WTS 60M) — 47.2 (2022-01-12) UniFormerV2-L — 47.8 (2022-09-22) UMT-L (ViT-L/16) — 48.7 (2023-03-28) OmniVec2 — 53.1 (2024-01-01)
RankModel Top 1 AccuracyTop 5 Accuracy Extra Training Data PaperCodeYear
1 OmniVec2 53.1– ✓ OmniVec2 - A Novel Transformer based Network for Large Scale Multimodal and Multitask Learning 2024
2 InternVideo2-1B 50.9– ✓ InternVideo2: Scaling Foundation Models for Multimodal Video Understanding opengvlab/internvideo · opengvlab/internvideo2 2024
3 UMT-L (ViT-L/16) 48.778.2 ✓ Unmasked Teacher: Towards Training-Efficient Video Foundation Models opengvlab/unmasked_teacher 2023
4 UniFormerV2-L 47.876.9 ✓ UniFormerV2: Spatiotemporal Learning by Arming Image ViTs with Video UniFormer OpenGVLab/UniFormerV2 · innat/UniFormerV2 2022
5 MTV-H (WTS 60M) 47.275.7 ✓ Multiview Transformers for Video Recognition google-research/scenic 2022
6 CoVeR(JFT-3B) 46.175.4 ✓ Co-training Transformer with Videos and Images Improves Action Recognition 2021
7 CoVeR(JFT-300M) 45.073.9 ✓ Co-training Transformer with Videos and Images Improves Action Recognition 2021
8 VATT-Large 41.167.7 ✓ VATT: Transformers for Multimodal Self-Supervised Learning from Raw Video, Audio and Text google-research/google-research · akashe/ProgrammingInterview · pwc-1/Paper-9 · +2 2021
9 MoViNet-A6 40.2– MoViNets: Mobile Video Networks for Efficient Video Recognition tensorflow/models · towhee-io/towhee · Atze00/MoViNet-pytorch 2021
10 MoViNet-A5 39.1– MoViNets: Mobile Video Networks for Efficient Video Recognition tensorflow/models · towhee-io/towhee · Atze00/MoViNet-pytorch 2021
11 MoViNet-A4 37.9– MoViNets: Mobile Video Networks for Efficient Video Recognition tensorflow/models · towhee-io/towhee · Atze00/MoViNet-pytorch 2021
12 VTN 37.465.4 ✓ Video Transformer Network bomri/SlowFast 2021
13 MBT (AV) 37.361.2 Attention Bottlenecks for Multimodal Fusion google-research/scenic 2021
14 MoViNet-A3 35.6– MoViNets: Mobile Video Networks for Efficient Video Recognition tensorflow/models · towhee-io/towhee · Atze00/MoViNet-pytorch 2021
15 MoViNet-A2 34.3– MoViNets: Mobile Video Networks for Efficient Video Recognition tensorflow/models · towhee-io/towhee · Atze00/MoViNet-pytorch 2021
16 AssembleNet 34.27%62.71% AssembleNet: Searching for Multi-Stream Neural Connectivity in Video Architectures tensorflow/models · google-research/google-research 2019
17 SRTG r3d-101 33.5658.49 Learn to cycle: Time-consistent feature discovery for action recognition alexandrosstergiou/Squeeze-and-Recursion-Temporal-Gates 2020
18 CoST (ResNet-101, 32 frames) 32.4%60.0% Collaborative Spatiotemporal Feature Learning for Video Action Recognition hikvision-research/cost 2019
19 MoViNet-A1 32.0– MoViNets: Mobile Video Networks for Efficient Video Recognition tensorflow/models · towhee-io/towhee · Atze00/MoViNet-pytorch 2021
20 EvaNet 31.8%– Evolving Space-Time Neural Architectures for Videos 2018
1–20 / 29 다음 → 페이지당 10 20 50 100