R2+1D-BERT

Reported on 4 benchmarks across 2 tasks · 1 paper · 2 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Robots2 results

Activity RecognitiononUCF 101
3-fold Accuracy· 2020-08-03
98.69
SOTA
Late Temporal Modeling in 3D CNN Architectures with BERT for Action Recognition arXiv:2008.01232
Activity RecognitiononHMDB-51
Average accuracy of 3 splits· uses extra data· 2020-08-03
85.1
best: 88.7 (VideoMAE V2-g)
Late Temporal Modeling in 3D CNN Architectures with BERT for Action Recognition arXiv:2008.01232

Time Series2 results

Action RecognitiononUCF 101
3-fold Accuracy· 2020-08-03
98.69
SOTA
Late Temporal Modeling in 3D CNN Architectures with BERT for Action Recognition arXiv:2008.01232
Action RecognitiononHMDB-51
Average accuracy of 3 splits· uses extra data· 2020-08-03
85.1
best: 88.7 (VideoMAE V2-g)
Late Temporal Modeling in 3D CNN Architectures with BERT for Action Recognition arXiv:2008.01232