Research map / Video / Video segmentation and tracking
Recurrent Residual Module for Fast Inference in Videos
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Apply RRM to AlexNet | AlexNet · Charades | speedup ratio: 146% | Helped |
| Apply RRM to AlexNet | AlexNet · UCF-101 | overall sparsity: 35.4% -> 60.1% | Helped |
| Apply RRM to Deep Compression | Deep Compression · MERL | speedup ratio: 210% | Helped |
| Apply RRM to XNOR-Net | XNOR-Net · MERL | speedup ratio: 927% | Helped |
| Apply RRM to rt-Pose | rt-Pose · MPII Video Pose | overall sparsity: 78.5% -> 91.0% | Helped |
| Apply RRM to YOLOv2 | YOLOv2 · Youtube-BB | mAP: 61.2 -> 61.1 | No clear effect |
| Apply RRM to rt-Pose | rt-Pose · MPII Video Pose | mAP: 46.2 -> 46.2 | No clear effect |
| Increase truncation threshold from 10^-2 to 10^-1 (ablation) | RRM with VGG-16 · UCF-101 | speedup ratio: 1.3× -> 1.8× | Helped |
| Increase truncation threshold (ablation) | RRM feature extraction · UCF-101 | top-1 precision | Hurt |
| Add accumulated error control scheme (ablation) | RRM · Video inference | precision | Helped |
Closest papers by meaning
- Deep Residual Learning for Image Recognition (CVPR 2016 · similarity 0.88)
- Spatiotemporal Residual Networks for Video Action Recognition (NeurIPS 2016 · similarity 0.87)
- LSTM Pose Machines (CVPR 2018 · similarity 0.87)
- FASTER Recurrent Networks for Efficient Video Classification (AAAI 2020 · similarity 0.87)
- VideoOneNet: Bidirectional Convolutional Recurrent OneNet with Trainable Data Steps for Video Processing (ICML 2020 · similarity 0.87)
- VA-RED$^2$: Video Adaptive Redundancy Reduction (ICLR 2021 · 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.