AI Paper Atlas

Research map / Video / Action recognition

Dynamic Network Quantization for Efficient Video Inference

Ximeng Sun, Rameswar Panda, Chun-Fu (Richard) Chen, Aude Oliva, Rogerio Feris, Kate Saenko

ICCV 2021 · Video › Action recognition · approach: Spatiotemporal CNNs for action recognition · 51 citations (OpenAlex)

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 changedWhereMeasuredEffect
add knowledge distillation loss from 32-bit model (ablation)VideoIQ · ActivityNetmAP: 73.5 -> 74.8Improved
add knowledge distillation loss from 32-bit model (ablation)VideoIQ · Mini-Sports1MmAP: 44.6 -> 46.4Improved
add efficiency loss L_e (ablation)VideoIQ · ActivityNetGFLOPs: 56.4 -> 34.6Improved
add efficiency loss L_e (ablation)VideoIQ · ActivityNetmAP: 75.1 -> 74.5Hurt
add regularizations L_b and L_d on top of L_e (ablation)VideoIQ · ActivityNetGFLOPs: 34.6 -> 28.1Improved
add L_b regularization alone (ablation)VideoIQ · ActivityNetmAP: 74.5 -> 74.3Hurt
add L_d regularization on top of L_b (ablation)VideoIQ · ActivityNetmAP: 74.3 -> 74.8Improved
add precision choices to skipping (decision space {32,4,2,0} vs {32,0}) (ablation)VideoIQ · ActivityNetmAP: 72.9 -> 74.8Improved
add skipping to precision choices (decision space {32,4,2,0} vs {32,4,2}) (ablation)VideoIQ · ActivityNetGFLOPs: 31.4 -> 28.1Improved
add skipping to precision choices (decision space {32,4,2,0} vs {32,4,2}) (ablation)VideoIQ · ActivityNetmAP: 74.5 -> 74.8Improved
use learned policy instead of random policy (ablation)VideoIQ · ActivityNetmAP: 72.8 -> 74.8Improved
add LSTM to the policy network (ablation)VideoIQ · ActivityNetmAP: 74.1 -> 74.8Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.

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

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy