AI Paper Atlas

Research map / Video / Video-language understanding

Weakly Supervised Summarization of Web Videos

Rameswar Panda, Abir Das, Ziyan Wu, Jan Ernst, Amit K. Roy-Chowdhury

ICCV 2017 · Video › Video-language understanding · approach: Dense video captioning · 103 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 cross-dataset pre-training on UCF101 instead of training from scratch (ablation)DeSumNet · CoSumtop-5 mAP relative to human (%): 71.4 -> 83.5Improved
add cross-dataset pre-training on UCF101 instead of training from scratch (ablation)DeSumNet · TVSumtop-5 mAP relative to human (%): 66.7 -> 75.2Improved
add refined web-video model adaptation to pre-training (ablation)DeSumNet · CoSumtop-5 mAP relative to human (%): 83.5 -> 87.7Improved
add refined web-video model adaptation to pre-training (ablation)DeSumNet · TVSumtop-5 mAP relative to human (%): 75.2 -> 80.8Improved
add enhanced data augmentation on top of pre-training and adaptation (ablation)DeSumNet · CoSumtop-5 mAP relative to human (%): 87.7 -> 88.5Improved
add enhanced data augmentation on top of pre-training and adaptation (ablation)DeSumNet · TVSumtop-5 mAP relative to human (%): 80.8 -> 82.2Improved
add unrefined noisy web videos directly to scratch training (ablation)DeSumNet · CoSumtop-5 mAP relative to human (%): 71.4 -> 76.3Improved
add unrefined noisy web videos directly to pre-trained training (ablation)DeSumNet · CoSumtop-5 mAP relative to human (%): 83.5 -> 84.4Improved
add unrefined noisy web videos directly to pre-trained training (ablation)DeSumNet · TVSumtop-5 mAP relative to human (%): 75.2 -> 77.3Improved
use refined web videos instead of unrefined noisy web videos after pre-training (ablation)DeSumNet · CoSumtop-5 mAP relative to human (%): 84.4 -> 87.7Improved
enforce diversity by clustering segments before summary selection (ablation)DeSumNet · CoSumtop-5 mAP: ~0.8% improvementNo clear effect
replace CVS with DeSumNet for time-lapsetime-lapse generation · CoSumaverage human rating: 3.23 -> 4.03Improved

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