Alert button
Picture for Jaehyun Jun

Jaehyun Jun

Alert button

Bilinear Attention Networks

Oct 19, 2018
Jin-Hwa Kim, Jaehyun Jun, Byoung-Tak Zhang

Figure 1 for Bilinear Attention Networks
Figure 2 for Bilinear Attention Networks
Figure 3 for Bilinear Attention Networks
Figure 4 for Bilinear Attention Networks

Attention networks in multimodal learning provide an efficient way to utilize given visual information selectively. However, the computational cost to learn attention distributions for every pair of multimodal input channels is prohibitively expensive. To solve this problem, co-attention builds two separate attention distributions for each modality neglecting the interaction between multimodal inputs. In this paper, we propose bilinear attention networks (BAN) that find bilinear attention distributions to utilize given vision-language information seamlessly. BAN considers bilinear interactions among two groups of input channels, while low-rank bilinear pooling extracts the joint representations for each pair of channels. Furthermore, we propose a variant of multimodal residual networks to exploit eight-attention maps of the BAN efficiently. We quantitatively and qualitatively evaluate our model on visual question answering (VQA 2.0) and Flickr30k Entities datasets, showing that BAN significantly outperforms previous methods and achieves new state-of-the-arts on both datasets.

* Accepted by NIPS 2018; Figure 1 was updated 
Viaarxiv icon

Overcoming Catastrophic Forgetting by Incremental Moment Matching

Jan 30, 2018
Sang-Woo Lee, Jin-Hwa Kim, Jaehyun Jun, Jung-Woo Ha, Byoung-Tak Zhang

Figure 1 for Overcoming Catastrophic Forgetting by Incremental Moment Matching
Figure 2 for Overcoming Catastrophic Forgetting by Incremental Moment Matching
Figure 3 for Overcoming Catastrophic Forgetting by Incremental Moment Matching
Figure 4 for Overcoming Catastrophic Forgetting by Incremental Moment Matching

Catastrophic forgetting is a problem of neural networks that loses the information of the first task after training the second task. Here, we propose a method, i.e. incremental moment matching (IMM), to resolve this problem. IMM incrementally matches the moment of the posterior distribution of the neural network which is trained on the first and the second task, respectively. To make the search space of posterior parameter smooth, the IMM procedure is complemented by various transfer learning techniques including weight transfer, L2-norm of the old and the new parameter, and a variant of dropout with the old parameter. We analyze our approach on a variety of datasets including the MNIST, CIFAR-10, Caltech-UCSD-Birds, and Lifelog datasets. The experimental results show that IMM achieves state-of-the-art performance by balancing the information between an old and a new network.

* Accepted for NIPS 2017 
Viaarxiv icon