AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Weakly-Supervised Crowd Counting Learns from Sorting rather than Locations

Yifan Yang, Guorong Li, Zhe Wu, Li Su, Qingming Huang, Nicu Sebe

ECCV 2020 · Label-efficient and robust learning › Recognition and long-tail · approach: Loss design and classifier theory · 91 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 the soft-sort network (sorting) to the counting network (ablation)regression network · WorldExpo10MAE: 20.1 -> 9.6Improved
add the regression network to the sorting network (ablation)sorting network · WorldExpo10sorting accuracy (%): 50.6 -> 78.2Improved
use Rayleigh soft-labels instead of hard one-hot labels (ablation)sorting network · WorldExpo10MAE: 13.2 -> 9.6Improved
use Rayleigh soft-labels instead of hard one-hot labels (ablation)sorting network · WorldExpo10sorting accuracy (%): 89.1 -> 78.2Hurt
replace VGG backbone with R3D 3D-conv backbone (ablation)full method · WorldExpo10MAE: 9.6 -> 10.1Hurt
use max and average pooling instead of only average pooling (ablation)pyramidal pooling · WorldExpo10MAE: 11.4 -> 9.6Improved
sort 4 frames instead of 3 (ablation)sorting network · WorldExpo10MAE: 9.6 -> 12.5Hurt
sort 5 frames instead of 3 (ablation)sorting network · WorldExpo10MAE: 9.6 -> 13.1Hurt
sort 4 frames instead of 3 (ablation)sorting network · WorldExpo10sorting accuracy (%): 78.2 -> 30.1Hurt
sort 5 frames instead of 3 (ablation)sorting network · WorldExpo10sorting accuracy (%): 78.2 -> 16.3Hurt
sort 5 frames instead of no sorting network (ablation)regression network · WorldExpo10MAE: 20.1 -> 13.1Improved

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