Outsource work to the Twine expert freelance network
![](https://assets-global.website-files.com/61a36f46798f2d52e14b678a/61a36f46798f2d6c664b689f_Vector-1.png)
support@twine.net
![](https://assets-global.website-files.com/61a36f46798f2d52e14b678a/61a36f46798f2deec94b689c_Vector.png)
+44-161-710-3084
Lipreading is the task of decoding text from the movement of a speaker’s mouth. Based on LipNet, a model that maps a variable-length sequence of video frames to text, making use of spatiotemporal convolutions, a recurrent network, and the connectionist temporal classification loss, trained entirely end-to-end.