1. Vision-Language models

2. Image captioning

3. Visual QA

4. Audio-visual speech recognition