Research map / Video / Action recognition
AR-Net: Adaptive Frame Resolution for Efficient Action Recognition
Yue Meng, Chung-Ching Lin, Rameswar Panda, Prasanna Sattigeri, Leonid Karlinsky, Aude Oliva, Kate Saenko, Rogerio Feris
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add frame skipping to the resolution-only action space (ablation) | AR-Net (ResNet) · ActivityNet-v1.3 | mAP: 73.4 -> 73.8 | Improved |
| add adaptive resolution to the skipping-only action space (ablation) | AR-Net (ResNet) · ActivityNet-v1.3 | mAP: 72.7 -> 73.8 | Improved |
| add efficiency (FLOPS) loss to accuracy loss (ablation) | AR-Net (ResNet) · ActivityNet-v1.3 | GFLOPS/V: 60.06 -> 36.48 | Improved |
| add efficiency (FLOPS) loss to accuracy loss (ablation) | AR-Net (ResNet) · ActivityNet-v1.3 | mAP: 74.5 -> 73.8 | Hurt |
| add uniform-usage loss regularizer (ablation) | AR-Net (ResNet) · ActivityNet-v1.3 | GFLOPS/V: 36.48 -> 33.47 | Improved |
| add uniform-usage loss regularizer (ablation) | AR-Net (ResNet) · ActivityNet-v1.3 | mAP: 73.8 -> 73.8 | No clear effect |
| add warm-up stage before joint training (ablation) | AR-Net (ResNet) · ActivityNet-v1.3 | mAP: 67.1 -> 73.3 | Improved |
| add fine-tuning stage after joint training (ablation) | AR-Net (ResNet) · ActivityNet-v1.3 | mAP: 73.3 -> 73.8 | Improved |
| replace random policy with learned policy | AR-Net (ResNet) · ActivityNet-v1.3 | mAP: 65.0 -> 73.8 | Improved |
| replace random policy with learned policy | AR-Net (ResNet) · FCVID | mAP: 75.3 -> 81.3 | Improved |
| replace uniform full-resolution processing with adaptive policy | ResNet-50 · ActivityNet-v1.3 | mAP: 72.5 -> 73.8 | Improved |
| replace Multi-Scale baseline with AR-Net | ResNet · ActivityNet-v1.3 | GFLOPS/V: 110.43 -> 33.47 | Improved |
Later papers that compared with this method
12 later accepted papers report AR-Net as a baseline in their results tables; 12 of them report a better number on most of the results they share with it.
- UCF-Crime-DVS: A Novel Event-Based Dataset for Video Anomaly Detection with Spiking Neural Networks (AAAI 2025 · better on 2 of 2 shared results)
- Facial Expression Recognition with Adaptive Frame Rate based on Multiple Testing Correction (ICML 2023 · better on 1 of 1 shared result)
- OCSampler: Compressing Videos to One Clip With Single-Step Sampling (CVPR 2022 · better on 1 of 1 shared result)
- Rethinking Resolution in the Context of Efficient Video Recognition (NeurIPS 2022 · better on 1 of 1 shared result)
- Delving into Details: Synopsis-to-Detail Networks for Video Recognition (ECCV 2022 · better on 1 of 1 shared result)
- NSNet: Non-Saliency Suppression Sampler for Efficient Video Recognition (ECCV 2022 · better on 1 of 1 shared result)
- Efficient Action Recognition via Dynamic Knowledge Propagation (ICCV 2021 · better on 7 of 7 shared results)
- FrameExit: Conditional Early Exiting for Efficient Video Recognition (CVPR 2021 · better on 4 of 4 shared results)
4 more not listed.
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- LiteEval — LiteEval: A Coarse-to-Fine Framework for Resource Efficient Video Recognition (NeurIPS 2019) · better on 1 of 1 shared result
- SCSampler — SCSampler: Sampling Salient Clips From Video for Efficient Action Recognition (ICCV 2019) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Look More but Care Less in Video Recognition (NeurIPS 2022 · similarity 0.88)
- EAN: Event Adaptive Network for Enhanced Action Recognition (IJCV 2022 · similarity 0.88)
- MGSampler: An Explainable Sampling Strategy for Video Action Recognition (ICCV 2021 · similarity 0.88)
- More Is Less: Learning Efficient Video Representations by Big-Little Network and Depthwise Temporal Aggregation (NeurIPS 2019 · similarity 0.88)
- Action Recognition With Spatial-Temporal Discriminative Filter Banks (ICCV 2019 · similarity 0.87)
- Optimizing Factorized Encoder Models: Time and Memory Reduction for Scalable and Efficient Action Recognition (ECCV 2024 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
