Compara mètodes

	Aprenentatge per Renforça Multimodal	Aprenentatge per Reforç Auto-supervisat
Camp	Aprenentatge profund	Aprenentatge profund
Família	Machine learning	Machine learning
Any d'origen	2015–2022	2020
Autor original	Multiple contributors (DeepMind, OpenAI, Google Brain, 2010s–2020s)	Laskin, M.; Srinivas, A.; Abbeel, P. (and contemporaries)
Tipus	Multimodal deep RL agent	Self-supervised auxiliary-task learning for RL
Font seminal	Reed, S., Zolna, K., Parisotto, E., Colmenarejo, S. G., Novikov, A., Barth-Maron, G., ... & de Freitas, N. (2022). A Generalist Agent. Transactions on Machine Learning Research.	Laskin, M., Srinivas, A., & Abbeel, P. (2020). CURL: Contrastive Unsupervised Representations for Reinforcement Learning. Proceedings of the 37th International Conference on Machine Learning (ICML), PMLR 119, 5639–5650.
Àlies	Multimodal RL, Multi-Sensory Reinforcement Learning, Vision-Language RL, Multi-Input RL	SSL-RL, self-supervised RL, representation-based reinforcement learning, auxiliary-task RL
Relacionats	6	4
Resum	Multimodal Reinforcement Learning trains agents to make sequential decisions by perceiving and integrating multiple input modalities — such as raw pixels, language instructions, audio, and proprioceptive sensors — simultaneously. Rather than acting on a single data stream, the agent fuses heterogeneous signals into a unified state representation and learns a policy through environmental reward feedback.	Self-supervised Reinforcement Learning (SSL-RL) augments standard RL training with self-supervised auxiliary objectives — such as contrastive, predictive, or data-augmentation-based tasks — applied to the agent's own experience. These objectives improve the quality of learned representations without requiring extra human labels, enabling faster convergence and better sample efficiency, especially in high-dimensional observation spaces like raw pixels.