Visual speech recognition (VSR) aims to recognise the content of speech based on the lip movements without relying on the audio stream. Advances in deep learning and the availability of large audio-visual datasets have led to the development of much more accurate and robust VSR models than ever before. However, these advances are usually due to larger training sets rather than the model design. In this work, we demonstrate that designing better models is equally important to using larger training sets. We propose the addition of prediction-based auxiliary tasks to a VSR model and highlight the importance of hyper-parameter optimisation and appropriate data augmentations. We show that such model works for different languages (English, Mandarin, Spanish, French, Portuguese and Italian) and outperforms all previous methods trained on publicly available datasets by a large margin. It even outperforms models that were trained on non-publicly available datasets containing up to to 21 times more data. We show furthermore that using additional training data, even in other languages or with automatically generated transcriptions, results in further improvement.
Google Cloud Console Incident #19008
We are currently experiencing an issue with authentication to Google App Engine sites, the Google Cloud Console, Identity Aware Proxy, and Google OAuth 2.0 endpoints.
Incident began at 2019-08-19 11:30 and ended at 2019-08-19 13:27 (all times are US/Pacific).
The Network Contagion Research Institute (NCRI) deploys machine learning tools to expose hate on digital social networks within a cross-platform, public-minded, and global framework. We are a multidisciplinary group of scientists and engineers who apply our technical skills to further public insight into the problem of online hate. We examine how hateful images and language grow within and between Web communities and how the infection of hate spreads between the online and the real world.
The condition of hate afflicts our capacity to see ourselves clearly, speak plainly to one another, and to assume collective and personal responsibility for our conditions.