OmniCount: Multi-label Object Counting with Semantic-Geometric Priors
Open paper
Benchmark
Model
Rank
Results
object-counting-on-fsc147
Omnicount (Open vocabulary, multi-label, without training)
–
MAE(test): 18.63
RMSE(test): 112
object-counting-on-pascal-voc-2007-count-test
Omnicount
–
mRMSE: 0.0023
mRMSE-nz: 0.009
Rank counts only results with a code link.