Supervising the Transfer of Reasoning Patterns in VQA
This addresses generalization issues in VQA by improving reasoning transfer, though it is incremental as it builds on existing methods for knowledge transfer.
The paper tackles the problem of transferring learned reasoning patterns from oracle-trained models to deployable VQA systems, proposing a regularization-based method that reduces sample complexity and shows effectiveness on the GQA dataset.
Methods for Visual Question Anwering (VQA) are notorious for leveraging dataset biases rather than performing reasoning, hindering generalization. It has been recently shown that better reasoning patterns emerge in attention layers of a state-of-the-art VQA model when they are trained on perfect (oracle) visual inputs. This provides evidence that deep neural networks can learn to reason when training conditions are favorable enough. However, transferring this learned knowledge to deployable models is a challenge, as much of it is lost during the transfer. We propose a method for knowledge transfer based on a regularization term in our loss function, supervising the sequence of required reasoning operations. We provide a theoretical analysis based on PAC-learning, showing that such program prediction can lead to decreased sample complexity under mild hypotheses. We also demonstrate the effectiveness of this approach experimentally on the GQA dataset and show its complementarity to BERT-like self-supervised pre-training.