Continual learning of conjugated visual representations through higher-order motion flows

• Published in: Neural networks Vol. 187; p. 107296
• Main Authors: Marullo, Simone, Tiezzi, Matteo, Gori, Marco, Melacci, Stefano
• Format: Journal Article
• Language: English
• Published: United States Elsevier Ltd 01.07.2025
• Subjects: Feature extraction from a single video stream; Humans; Learning from constraints; Lifelong learning; Machine Learning; Motion; Neural Networks, Computer; Online learning; Learning from constraints; Lifelong learning; Feature extraction from a single video stream; Online learning

Learning with neural networks from a continuous stream of visual information presents several challenges due to the non-i.i.d. nature of the data. However, it also offers novel opportunities to develop representations that are consistent with the information flow. In this paper we investigate the ca...