Ask, Attend and Answer: Exploring Question-Guided Spatial Attention for Visual Question Answering

2015/11/17 by Huijuan Xu, Kate Saenko, Xu, Huijuan +1 · 105 citations
Computer Science · #Advanced Image and Video Retrieval Techniques #Artificial intelligence #Closed captioning #Computer science #Convolutional neural network #Domain Adaptation and Few-Shot Learning #Image (mathematics) #Inference #Machine learning #Multimodal Machine Learning Applications #Natural language processing #Process (computing) #Question answering #Task (project management) #cs.AI #cs.CL #cs.CV #cs.NE

paper · pdf · doi:10.48550/arxiv.1511.05234

published in arXiv (Cornell University) (Cornell University) · include test-standard result on VQA full release (V1.0) dataset

openalex publication_date 2015/11/17 · arxiv created 2016/03/19 · arxiv updated 2016/03/22 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/06

Abstract

We address the problem of Visual Question Answering (VQA), which requires joint image and language understanding to answer a question about a given photograph. Recent approaches have applied deep image captioning methods based on convolutional-recurrent networks to this problem, but have failed to model spatial inference. To remedy this, we propose a model we call the Spatial Memory Network and apply it to the VQA task. Memory networks are recurrent neural networks with an explicit attention mechanism that selects certain parts of the information stored in memory. Our Spatial Memory Network stores neuron activations from different spatial regions of the image in its memory, and uses the question to choose relevant regions for computing the answer, a process of which constitutes a single "hop" in the network. We propose a novel spatial attention architecture that aligns words with image patches in the first hop, and obtain improved results by adding a second attention hop which considers the whole question to choose visual evidence based on the results of the first hop. To better understand the inference process learned by the network, we design synthetic questions that specifically require spatial inference and visualize the attention weights. We evaluate our model on two published visual question answering datasets, DAQUAR [1] and VQA [2], and obtain improved results compared to a strong deep baseline model (iBOWIMG) which concatenates image and question features to predict the answer [3].

Citations

• VQA: Visual Question Answering
• Ask Your Neurons: A Neural-based Approach to Answering Questions about Images

Cited by

• BiCoRec: Bias-Mitigated Context-Aware Sequential Recommendation Model
• An Analysis of Visual Question Answering Algorithms
• A Focused Dynamic Attention Model for Visual Question Answering
• Sparsely Grouped Multi-task Generative Adversarial Networks for Facial Attribute Manipulation
• Visual Question Answering: A Survey of Methods and Datasets
• Motion-Appearance Co-Memory Networks for Video Question Answering
• A Simple Loss Function for Improving the Convergence and Accuracy of Visual Question Answering Models
• HydraPlus-Net: Attentive Deep Features for Pedestrian Analysis
• Simple Baseline for Visual Question Answering
• Factor Graph Attention
• Understanding Neural Networks through Representation Erasure
• Traversing the Continuous Spectrum of Image Retrieval with Deep Dynamic Models
• Deep Multi-Modal Sets
• Transparency by Design: Closing the Gap Between Performance and Interpretability in Visual Reasoning
• Attend in groups: a weakly-supervised deep learning framework for learning from web data
• Aspect-augmented Adversarial Networks for Domain Adaptation
• Multi-modal Factorized Bilinear Pooling with Co-Attention Learning for Visual Question Answering
• Hierarchical Question-Image Co-Attention for Visual Question Answering
• Modulating early visual processing by language
• Incorporating External Knowledge to Answer Open-Domain Visual Questions with Dynamic Memory Networks
• Progressive Attention Networks for Visual Attribute Prediction
• Segmentation from Natural Language Expressions
• Show, Tell and Summarize: Dense Video Captioning Using Visual Cue Aided Sentence Summarization
• An Empirical Evaluation of Visual Question Answering for Novel Objects
• Implicit Distortion and Fertility Models for Attention-based Encoder-Decoder NMT Model
• Multi-step Joint-Modality Attention Network for Scene-Aware Dialogue System
• Creativity: Generating Diverse Questions using Variational Autoencoders
• Neural Attention Models for Sequence Classification: Analysis and Application to Key Term Extraction and Dialogue Act Detection
• Tell-and-Answer: Towards Explainable Visual Question Answering using Attributes and Captions
• Learning Visual Knowledge Memory Networks for Visual Question Answering
• Bottom-Up and Top-Down Attention for Image Captioning and Visual Question Answering
• MarioQA: Answering Questions by Watching Gameplay Videos
• Learn to Combine Modalities in Multimodal Deep Learning
• Learning to Disambiguate by Asking Discriminative Questions

Related

• Hierarchical Question-Image Co-Attention for Visual Question Answering
• Visual Question Answering Using Semantic Information from Image Descriptions
• ABC-CNN: An Attention Based Convolutional Neural Network for Visual Question Answering
• Stacked Attention Networks for Image Question Answering
• From Pixels to Objects: Cubic Visual Attention for Visual Question Answering