Crowd Counting On Shanghaitech A
Metrics
MAE
Results
Performance results of various models on this benchmark
Comparison Table
| Model Name | MAE |
|---|---|
| scale-aggregation-network-for-accurate-and | 67.0 |
| cnn-based-cascaded-multi-task-learning-of | 101.3 |
| improving-deep-regression-with-ordinal | 65.6 |
| segmentation-guided-attention-network-for | 57.6 |
| context-aware-crowd-counting | 62.3 |
| rethinking-spatial-invariance-of-1 | 54.8 |
| clip-ebc-clip-can-count-accurately-through | 66.3 |
| encoder-decoder-based-convolutional-neural | 57.55 |
| from-open-set-to-closed-set-counting-objects | 58.3 |
| csrnet-dilated-convolutional-neural-networks | 68.2 |
| divide-and-grow-capturing-huge-diversity-in | 72.5 |
| crowd-counting-with-deep-negative-correlation | 73.5 |
| generating-high-quality-crowd-density-maps | 73.6 |
| clip-ebc-clip-can-count-accurately-through | 52.5 |
| clip-ebc-clip-can-count-accurately-through | 62.3 |
| improving-local-features-with-relevant | 54.8 |
| leveraging-unlabeled-data-for-crowd-counting | 73.6 |
| clip-ebc-clip-can-count-accurately-through | 54.0 |
| crowd-counting-and-individual-localization | 49.9 |
| cross-scene-crowd-counting-via-deep | 181.8 |
| iterative-correlation-based-feature | 73.70 |
| distribution-matching-for-crowd-counting | 59.7 |
| improving-point-based-crowd-counting-and | 48.8 |
| fusioncount-efficient-crowd-counting-via | 62.2 |
| crowd-counting-via-adversarial-cross-scale | 75.7 |
| segmentation-guided-attention-network-for | 58 |
| learning-spatial-awareness-to-improve-crowd | 59.4 |
| iterative-crowd-counting | 68.5 |
| single-image-crowd-counting-via-multi-column-1 | 110.2 |
| fgenet-fine-grained-extraction-network-for-2 | 51.66 |
| locate-size-and-count-accurately-resolving | 66.4 |
| switching-convolutional-neural-network-for | 90.4 |
| rethinking-counting-and-localization-in | 52.74 |
| vmambacc-a-visual-state-space-model-for-crowd | 51.87 |