ys-zong/awesome-self-supervised-multimodal-learning

This resource curates the latest research in self-supervised multimodal learning, helping researchers and machine learning engineers develop models that understand information from various data types like images, text, and audio, without needing extensive human-labeled data. It provides a structured list of papers and code, categorizing approaches by objective and application. The target audience is academic researchers and advanced practitioners in AI/ML looking to build sophisticated, data-efficient multimodal AI systems.

277 stars. No commits in the last 6 months.