Abstract | ||
---|---|---|
The goal of this work is to recognise phrases and sentences being spoken by a talking face, with or without the audio. Unlike previous works that have focussed on recognising a limited number of words or phrases, we tackle lip reading as an
<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">open-world</i>
problem – unconstrained natural language sentences, and in the wild videos. Our key contributions are: (1) we compare two models for lip reading, one using a CTC loss, and the other using a sequence-to-sequence loss. Both models are built on top of the transformer self-attention architecture; (2) we investigate to what extent lip reading is complementary to audio speech recognition, especially when the audio signal is noisy; (3) we introduce and publicly release a new dataset for audio-visual speech recognition, LRS2-BBC, consisting of thousands of natural sentences from British television. The models that we train surpass the performance of all previous work on a lip reading benchmark dataset by a significant margin. |
Year | DOI | Venue |
---|---|---|
2018 | 10.1109/TPAMI.2018.2889052 | IEEE Transactions on Pattern Analysis and Machine Intelligence |
Keywords | Field | DocType |
Humans,Speech Perception,Algorithms,Lipreading,Speech | Computer vision,Audio signal,Architecture,Computer science,Speech recognition,Natural language,Audio-visual speech recognition,Artificial intelligence | Journal |
Volume | Issue | ISSN |
44 | 12 | 0162-8828 |
Citations | PageRank | References |
22 | 0.68 | 34 |
Authors | ||
5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Triantafyllos Afouras | 1 | 121 | 9.19 |
Joon Son Chung | 2 | 208 | 20.20 |
Andrew Senior | 3 | 4687 | 260.55 |
Oriol Vinyals | 4 | 9419 | 418.45 |
Andrew Zisserman | 5 | 45998 | 3200.71 |