Papers › MRCNet: Crowd Counting and Density Map Estimation in Aerial and Ground Imagery

MRCNet: Crowd Counting and Density Map Estimation in Aerial and Ground Imagery

27 Sep 2019arXiv:1909.12743archive 2025-07-28

Reza Bahmanyar, Elenora Vig, Peter Reinartz

In spite of the many advantages of aerial imagery for crowd monitoring and management at mass events, datasets of aerial images of crowds are still lacking in the field. As a remedy, in this work we introduce a novel crowd dataset, the DLR Aerial Crowd Dataset (DLR-ACD), which is composed of 33 large aerial images acquired from 16 flight campaigns over mass events with 226,291 persons annotated. To the best of our knowledge, DLR-ACD is the first aerial crowd dataset and will be released publicly. To tackle the problem of accurate crowd counting and density map estimation in aerial images of crowds, this work also proposes a new encoder-decoder convolutional neural network, the so-called Multi-Resolution Crowd Network MRCNet. The encoder is based on the VGG-16 network and the decoder is composed of a set of bilinear upsampling and convolutional layers. Using two losses, one at an earlier level and another at the last level of the decoder, MRCNet estimates crowd counts and high-resolution crowd density maps as two different but interrelated tasks. In addition, MRCNet utilizes contextual and detailed local information by combining high- and low-level features through a number of lateral connections inspired by the Feature Pyramid Network (FPN) technique. We evaluated MRCNet on the proposed DLR-ACD dataset as well as on the ShanghaiTech dataset, a CCTV-based crowd counting benchmark. The results demonstrate that MRCNet outperforms the state-of-the-art crowd counting methods in estimating the crowd counts and density maps for both aerial and CCTV-based images.

PaperPDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

gpspelle/Crowd-Counting mentioned on GitHubMIT 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

Crowd CountingDecoderManagement

Datasets

Introduced by this paper, per the archive.

DLR-ACD

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Crowd Counting DLR-ACD MRCNet (ours) F1-score 0.49 #1 of 5 Archive leaderboard report
Crowd Counting DLR-ACD MRCNet (ours) MAE 906 #1 of 5 Archive leaderboard report
Crowd Counting DLR-ACD MRCNet (ours) MNAE 0.21 #1 of 5 Archive leaderboard report
Crowd Counting DLR-ACD MRCNet (ours) Precision 0.51 #1 of 5 Archive leaderboard report
Crowd Counting DLR-ACD MRCNet (ours) RMSE 1307.4 #1 of 5 Archive leaderboard report
Crowd Counting DLR-ACD MRCNet (ours) Recall 0.48 #1 of 5 Archive leaderboard report
Crowd Counting DLR-ACD ic-CNN F1-score 0.46 #2 of 5 Archive leaderboard report
Crowd Counting DLR-ACD ic-CNN MAE 1481.3 #2 of 5 Archive leaderboard report
Crowd Counting DLR-ACD ic-CNN MNAE 0.72 #2 of 5 Archive leaderboard report
Crowd Counting DLR-ACD ic-CNN Precision 0.44 #2 of 5 Archive leaderboard report
Crowd Counting DLR-ACD ic-CNN RMSE 2087 #2 of 5 Archive leaderboard report
Crowd Counting DLR-ACD ic-CNN Recall 0.52 #2 of 5 Archive leaderboard report
Crowd Counting DLR-ACD Liu et al F1-score 0.44 #3 of 5 Archive leaderboard report
Crowd Counting DLR-ACD Liu et al MAE 833.3 #3 of 5 Archive leaderboard report
Crowd Counting DLR-ACD Liu et al MNAE 0.25 #3 of 5 Archive leaderboard report
Crowd Counting DLR-ACD Liu et al Precision 45 #3 of 5 Archive leaderboard report
Crowd Counting DLR-ACD Liu et al RMSE 1085.9 #3 of 5 Archive leaderboard report
Crowd Counting DLR-ACD Liu et al Recall 0.44 #3 of 5 Archive leaderboard report
Crowd Counting DLR-ACD MCNN F1-score 0.39 #4 of 5 Archive leaderboard report
Crowd Counting DLR-ACD MCNN MAE 1989.7 #4 of 5 Archive leaderboard report
Crowd Counting DLR-ACD MCNN MNAE 0.87 #4 of 5 Archive leaderboard report
Crowd Counting DLR-ACD MCNN Precision 0.43 #4 of 5 Archive leaderboard report
Crowd Counting DLR-ACD MCNN RMSE 3016.3 #4 of 5 Archive leaderboard report
Crowd Counting DLR-ACD MCNN Recall 0.41 #4 of 5 Archive leaderboard report
Crowd Counting DLR-ACD CSRNet F1-score 0.24 #5 of 5 Archive leaderboard report
Crowd Counting DLR-ACD CSRNet MAE 3388.8 #5 of 5 Archive leaderboard report
Crowd Counting DLR-ACD CSRNet MNAE 0.71 #5 of 5 Archive leaderboard report
Crowd Counting DLR-ACD CSRNet Precision 0.2 #5 of 5 Archive leaderboard report
Crowd Counting DLR-ACD CSRNet RMSE 4456.5 #5 of 5 Archive leaderboard report
Crowd Counting DLR-ACD CSRNet Recall 0.33 #5 of 5 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.

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