Association: Remind Your GAN not to Forget
This work tackles catastrophic forgetting in GANs, a problem for researchers developing continually learning generative models, offering an incremental solution.
This paper addresses catastrophic forgetting in GANs, where models fail to retain old knowledge when learning new tasks. The authors propose a brain-inspired associative memory system that uses a heuristic mechanism to recall historical episodes and a distillation measure to dampen feature reconstruction for new tasks, demonstrating effectiveness on image-to-image translation.
Neural networks are susceptible to catastrophic forgetting. They fail to preserve previously acquired knowledge when adapting to new tasks. Inspired by human associative memory system, we propose a brain-like approach that imitates the associative learning process to achieve continual learning. We design a heuristics mechanism to potentiatively stimulate the model, which guides the model to recall the historical episodes based on the current circumstance and obtained association experience. Besides, a distillation measure is added to depressively alter the efficacy of synaptic transmission, which dampens the feature reconstruction learning for new task. The framework is mediated by potentiation and depression stimulation that play opposing roles in directing synaptic and behavioral plasticity. It requires no access to the original data and is more similar to human cognitive process. Experiments demonstrate the effectiveness of our method in alleviating catastrophic forgetting on image-to-image translation tasks.