Training Recurrent Answering Units with Joint Loss Minimization for VQA

TL;DR

This paper introduces a recurrent neural network for visual question answering that employs joint loss minimization across multiple units, enabling adaptive multi-step reasoning and improved accuracy over existing attention-based methods.

Contribution

It presents a novel joint loss training approach for recurrent answering units with early stopping, allowing flexible multi-step inference in VQA tasks.

Findings

Outperforms existing multi-step attention models in VQA accuracy

Effectively learns adaptive reasoning steps through early stopping

Shared parameters enable efficient multi-step inference

Abstract

We propose a novel algorithm for visual question answering based on a recurrent deep neural network, where every module in the network corresponds to a complete answering unit with attention mechanism by itself. The network is optimized by minimizing loss aggregated from all the units, which share model parameters while receiving different information to compute attention probability. For training, our model attends to a region within image feature map, updates its memory based on the question and attended image feature, and answers the question based on its memory state. This procedure is performed to compute loss in each step. The motivation of this approach is our observation that multi-step inferences are often required to answer questions while each problem may have a unique desirable number of steps, which is difficult to identify in practice. Hence, we always make the first unit in the network solve problems, but allow it to learn the knowledge from the rest of units by backpropagation unless it degrades the model. To implement this idea, we early-stop training each unit as soon as it starts to overfit. Note that, since more complex models tend to overfit on easier questions quickly, the last answering unit in the unfolded recurrent neural network is typically killed first while the first one remains last. We make a single-step prediction for a new question using the shared model. This strategy works better than the other options within our framework since the selected model is trained effectively from all units without overfitting. The proposed algorithm outperforms other multi-step attention based approaches using a single step prediction in VQA dataset.