AI Paper Atlas

Research map / Video / Video segmentation and tracking

How Severe Is Benchmark-Sensitivity in Video Self-Supervised Learning?

Fida Mohammad Thoker, Hazel Doughty, Piyush Bagad, Cees G. M. Snoek

ECCV 2022 · Video › Video segmentation and tracking · approach: Self-supervised video representation learning · 19 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 self-supervised video pre-training before finetuning, on UCF-101R(2+1)D-18 · UCF-101 (full finetuning)top-1 accuracy: 77.3 -> 91.3 (GDT)Improved
use image-based MoCo pre-training instead of no pre-training on SS-v2 and EK-100R(2+1)D-18 · SS-v2 / EPIC-Kitchens-100 (finetuning)top-1 accuracy: 57.1 -> 57.1 (SS-v2), 25.7 -> 26.4 (EK-100)No clear effect
use supervised Kinetics pre-training instead of the best self-supervised methodR(2+1)D-18 · UCF-101 (full finetuning)top-1 accuracy: 91.3 -> 93.9Improved
use supervised pre-training instead of the best self-supervised method under a large domain shiftR(2+1)D-18 · EPIC-Kitchens-100 (full finetuning)top-1 accuracy: 42.8 -> 47.7Improved
use supervised pre-training instead of best self-supervised for linear evaluationR(2+1)D-18 · Kinetics-400 (linear evaluation)top-1 accuracy: 46.0 -> 65.9Improved
use contrastive AVID-CMA instead of supervised pre-training in linear evaluation on NTU-60R(2+1)D-18 · NTU-60 (linear evaluation)top-1 accuracy: 45.5 -> 53.9Improved
reduce finetuning samples to 1000 on UCF-101self-supervised video methods · UCF-101 (low data)top-1 accuracyHurt
use CtP instead of the second-best self-supervised method on FX-S1R(2+1)D-18 · FineGym FX-S1mean per-class accuracy: 68.0 -> 79.1 (12%)Improved
use MoCo instead of VideoMoCo (frame dropout invariance) on fine-grained Gym-99 subsetsR(2+1)D-18 · FineGym FX-S1mean per-class accuracy: 57.3 -> 65.0Improved
switch from balanced Gym-99 to long-tailed Gym-288 for supervised pre-trainingR(2+1)D-18 · FineGymtop-1 accuracy: 88.6 -> 58.4Hurt
add self-supervised pre-training for spatio-temporal action detection on UCF-101R(2+1)D-18 · UCF-101 action detectionmAP: 0.327 -> 0.476 (TCLR)Improved
use supervised pre-training instead of self-supervised on Charades multi-label classificationR(2+1)D-18 · CharadesmAP: 12.2 -> 23.5Improved

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