3D Conv + ResNet-18 + Bi-GRU(Face Cutout)

Reported on 2 benchmarks across 2 tasks · 1 paper

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

Computer Vision1 result

LipreadingonLip Reading in the Wild
Top-1 Accuracy· 2020-03-06
85.02
best: 95 (SyncVSR (Word Boundary))
Can We Read Speech Beyond the Lips? Rethinking RoI Selection for Deep Visual Speech Recognition arXiv:2003.03206

Natural Language Processing1 result

Natural Language TransductiononLip Reading in the Wild
Top-1 Accuracy· 2020-03-06
85.02
best: 95 (SyncVSR (Word Boundary))
Can We Read Speech Beyond the Lips? Rethinking RoI Selection for Deep Visual Speech Recognition arXiv:2003.03206