AI Paper Atlas

Research map / Video / Action recognition

Massively Parallel Video Networks

Joao Carreira, Viorica Patraucean, Laurent Mazare, Andrew Zisserman, Simon Osindero

ECCV 2018 · Video › Action recognition · approach: Spatiotemporal CNNs for action recognition · 39 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
distill activations from a sequential teacher into the parallel model (ablation)Par-Inception parallel (10 subnets) · miniKinetics action recognitiontop-1 accuracy: 54.5 -> 61.2Improved
distill activations from a sequential teacher (ablation)Par-DenseNet with clocks · miniKinetics keypoint localisationweighted sigmoid cross-entropyImproved
increase parallelism from sequential to 10 subnets (ablation)Par-Inception · miniKinetics action recognitiontop-1 accuracy: 71.4 -> 54.5Hurt
increase parallelism from sequential to 14 subnets (ablation)Par-DenseNet · miniKinetics action recognitiontop-1 accuracy: 67.6 -> 54.0Hurt
add multi-rate clocks to sequential model (ablation)Par-DenseNet · Throughput 8 GPUsspeedup: 1.0 -> 3.4Improved
add multi-rate clocks to fully parallel model (ablation)Par-DenseNet · Throughput 8 GPUsspeedup: 3.8 -> 7.4Improved
add multi-rate clocks to fully parallel model (ablation)Par-Inception · Throughput 8 GPUsspeedup: 2.6 -> 5.0Improved
add multi-rate clocks (ablation)Par-DenseNet · miniKinetics keypoint localisationweighted sigmoid cross-entropyHurt
add temporal filters (ablation)Par-DenseNet · miniKinetics keypoint localisationweighted sigmoid cross-entropyMixed
add feedback connection (ablation)Par-DenseNet · miniKinetics keypoint localisationweighted sigmoid cross-entropyImproved
train with depth-parallel connectivity instead of rewiring a sequential model (ablation)Par-DenseNet · miniKinetics keypoint localisationtest lossImproved
allow prediction latency of 2 frames (ablation)Par-DenseNet up to 4 subnets · miniKinetics keypoint localisationtest lossImproved

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