AI Paper Atlas

Research map / Video / Action recognition

DistInit: Learning Video Representations Without a Single Labeled Video

Rohit Girdhar, Du Tran, Lorenzo Torresani, Deva Ramanan

ICCV 2019 · Video › Action recognition · approach: Contrastive video representation learning · 57 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
initialize with DistInit instead of scratchRes3D-18 · HMDB-51 split 1top-1 accuracy: 25.4 -> 39.9Improved
initialize with DistInit instead of ImageNet inflationRes3D-18 · HMDB-51 split 1top-1 accuracy: 35.8 -> 39.9Improved
initialize with DistInit instead of scratchR(2+1)D-18 · HMDB-51 split 1top-1 accuracy: 24.1 -> 40.3Improved
pick a random frame for teacher target instead of the center frame (ablation)DistInit R(2+1)D-18 · HMDB-51 split 1top-1 accuracy: 40.3 -> 43.2Improved
average teacher logits over 2 random frames instead of the center frame (ablation)DistInit R(2+1)D-18 · HMDB-51 split 1top-1 accuracy: 40.3 -> 44.0Improved
use cross-entropy over softmax instead of MSE over logits (ablation)DistInit R(2+1)D-18 · HMDB-51 split 1top-1 accuracy: 39.9 -> 40.3No clear effect
use Kinetics instead of Sports1M unlabeled videos (ablation)DistInit R(2+1)D-18 · HMDB-51 split 1top-1 accuracy: 39.9 -> 40.3No clear effect
combine Kinetics and Sports1M unlabeled videos (ablation)DistInit R(2+1)D-18 · HMDB-51 split 1top-1 accuracy: 40.3 -> 41.8Improved
use ImageNet teacher instead of PlaceNet teacher (ablation)DistInit R(2+1)D-18 · HMDB-51 split 1top-1 accuracy: 36.8 -> 40.3Improved
add PlaceNet teacher to the ImageNet teacher (ablation)DistInit R(2+1)D-18 · HMDB-51 split 1top-1 accuracy: 40.3 -> 40.7No clear effect
drop teacher low-confidence videos (entropy above 2.7, about 50% of data) (ablation)DistInit R(2+1)D-18 · HMDB-51 split 1top-1 accuracyImproved
initialize flow network with DistInit instead of ImageNet mean inflationRes3D-18 / R(2+1)D-18 on optical flow · HMDB-51 split 1top-1 accuracy: 43.9 -> 49.2Improved

Later papers that compared with this method

1 later accepted paper reports DistInit as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

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