AI Paper Atlas

Research map / Video / Action recognition

Memory-augmented Dense Predictive Coding for Video Representation Learning

Tengda Han, Weidi Xie, Andrew Zisserman

ECCV 2020 · Video › Action recognition · approach: Contrastive video representation learning · 217 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 compressive memory (size 1024) to DPC (ablation)DPC (R18, RGB) · UCF101 finetunetop-1 accuracy: 63.6 -> 68.2Improved
self-supervised MemDPC pretraining instead of random init (RGB) (ablation)R18 RGB · UCF101 finetunetop-1 accuracy: 61.8 -> 68.2Improved
self-supervised MemDPC pretraining instead of random init (Flow) (ablation)R18 Flow · UCF101 finetunetop-1 accuracy: 74.6 -> 81.9Improved
increase memory size from 512 to 1024 (ablation)MemDPC R18 RGB · UCF101 finetunetop-1 accuracy: 65.3 -> 68.2Improved
increase memory size from 1024 to 2048 (ablation)MemDPC R18 RGB · UCF101 finetunetop-1 accuracy: 68.2 -> 68.0No clear effect
add bi-directional aggregation (RGB) (ablation)MemDPC R18 RGB · UCF101 finetunetop-1 accuracy: 68.2 -> 69.2Improved
add bi-directional aggregation (Flow) (ablation)MemDPC R18 Flow · UCF101 finetunetop-1 accuracy: 81.9 -> 82.3Improved
add bi-directional aggregation (RGB+Flow) (ablation)MemDPC R18x2 · UCF101 finetunetop-1 accuracy: 84.0 -> 84.3Improved
fuse RGB and Flow streams instead of Flow only (ablation)MemDPC · UCF101 finetunetop-1 accuracy: 81.9 -> 84.0Improved
use non-linear probe instead of linear probe on frozen features (UCF101)MemDPC two-stream K400 · UCF101top-1 accuracy: 54.1 -> 58.5Improved
use non-linear probe instead of linear probe on frozen features (HMDB51)MemDPC two-stream K400 · HMDB51top-1 accuracy: 30.5 -> 33.6Improved
add two-stream fusion of RGB and Flow predictionsMemDPC K400 finetune · UCF101top-1 accuracy: 78.1 -> 86.1Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 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