Disentanglement of Latent Representations via Causal Interventions
This addresses the challenge of interpretable and controllable image generation for machine learning researchers, though it is incremental as it builds on existing disentanglement and causal representation learning approaches.
The paper tackled the problem of disentangling independent factors of variation in image generation by introducing a method that uses causal interventions on a graph of quantized vectors, showing it can effectively disentangle factors and perform precise interventions on semantic attributes without quality loss, even with imbalanced data.
The process of generating data such as images is controlled by independent and unknown factors of variation. The retrieval of these variables has been studied extensively in the disentanglement, causal representation learning, and independent component analysis fields. Recently, approaches merging these domains together have shown great success. Instead of directly representing the factors of variation, the problem of disentanglement can be seen as finding the interventions on one image that yield a change to a single factor. Following this assumption, we introduce a new method for disentanglement inspired by causal dynamics that combines causality theory with vector-quantized variational autoencoders. Our model considers the quantized vectors as causal variables and links them in a causal graph. It performs causal interventions on the graph and generates atomic transitions affecting a unique factor of variation in the image. We also introduce a new task of action retrieval that consists of finding the action responsible for the transition between two images. We test our method on standard synthetic and real-world disentanglement datasets. We show that it can effectively disentangle the factors of variation and perform precise interventions on high-level semantic attributes of an image without affecting its quality, even with imbalanced data distributions.