Papers › Learning Efficient Single-stage Pedestrian Detectors by Asymptotic Localization Fitting

Learning Efficient Single-stage Pedestrian Detectors by Asymptotic Localization Fitting

1 Sep 2018ECCV 2018 9archive 2025-07-28

Wei Liu, Shengcai Liao, Weidong Hu, Xuezhi Liang, Xiao Chen

Though Faster R-CNN based two-stage detectors have witnessed significant boost in pedestrian detection accuracy, it is still slow for practical applications. One solution is to simplify this working flow as a single-stage detector. However, current single-stage detectors (e.g. SSD) have not presented competitive accuracy on common pedestrian detection benchmarks. This paper is towards a successful pedestrian detector enjoying the speed of SSD while maintaining the accuracy of Faster R-CNN. Specifically, a structurally simple but effective module called emph{Asymptotic Localization Fitting} (ALF) is proposed, which stacks a series of predictors to directly evolve the default anchor boxes of SSD step by step into improving detection results. As a result, during training the latter predictors enjoy more and better-quality positive samples, meanwhile harder negatives could be mined with increasing IoU thresholds. On top of this, an efficient single-stage pedestrian detection architecture (denoted as ALFNet) is designed, achieving state-of-the-art performance on CityPersons and Caltech, two of the largest pedestrian detection benchmarks, and hence resulting in an attractive pedestrian detector in both accuracy and speed. Code is available at href{https://github.com/VideoObjectSearch/ALFNet}{https://github.com/VideoObjectSearch/ALFNet}.

PaperPDFCode

Code

VideoObjectSearch/ALFNet officialmentioned in papertf report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Pedestrian Detection

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Pedestrian Detection Caltech ALFNet + CityPersons dataset Reasonable Miss Rate 4.5 #12 of 33 Archive leaderboard report
Pedestrian Detection Caltech ALFNet Reasonable Miss Rate 6.1 #17 of 33 Archive leaderboard report
Pedestrian Detection CityPersons ALFNet Bare MR^-2 8.4 #15 of 22 Archive leaderboard report
Pedestrian Detection CityPersons ALFNet Heavy MR^-2 51.9 #15 of 22 Archive leaderboard report
Pedestrian Detection CityPersons ALFNet Large MR^-2 6.6 #15 of 22 Archive leaderboard report
Pedestrian Detection CityPersons ALFNet Medium MR^-2 5.7 #15 of 22 Archive leaderboard report
Pedestrian Detection CityPersons ALFNet Partial MR^-2 11.4 #15 of 22 Archive leaderboard report
Pedestrian Detection CityPersons ALFNet Reasonable MR^-2 12.0 #15 of 22 Archive leaderboard report
Pedestrian Detection CityPersons ALFNet Small MR^-2 19.0 #15 of 22 Archive leaderboard report
Pedestrian Detection CityPersons ALFNet Test Time 0.27 #15 of 22 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

1x1 ConvolutionConvolutionFaster R-CNNNon Maximum SuppressionRPNRoIPoolSPEEDSSDSoftmax

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections