A Generic Visualization Approach For Convolutional Neural Networks

Abstract

Retrieval networks are essential for searching and indexing. Compared to classification networks, attention visualization for retrieval networks is hardly studied. We formulate attention visualization as a constrained optimization problem. We leverage the unit L2-Norm constraint as an attention filter (L2-CAF) to localize attention in both classification and retrieval networks. Unlike recent literature, our approach requires neither architectural changes nor fine-tuning. Thus, a pre-trained network's performance is never undermined L2-CAF is quantitatively evaluated using weakly supervised object localization. State-of-the-art results are achieved on classification networks. For retrieval networks, significant improvement margins are achieved over a Grad-CAM baseline. Qualitative evaluation demonstrates how the L2-CAF visualizes attention per frame for a recurrent retrieval network. Further ablation studies highlight the computational cost of our approach and compare L2-CAF with othe

Stats

• citations6
• S2 citations—
• github stars0
• HF likes0
• heat score6.34
• arxiv keytaha2020a

Related papers

• Class-weighted Convolutional Features For Visual Instance Search (2017)12.81
• All The Attention You Need: Global-local, Spatial-channel Attention For Image Retrieval (2021)13.97
• Context Aware Query Image Representation For Particular Object Retrieval (2017)11.49
• Clothing Retrieval With Visual Attention Model (2017)12.10
• Channel Recurrent Attention Networks For Video Pedestrian Retrieval (2020)0.00
• DALG: Deep Attentive Local And Global Modeling For Image Retrieval (2022)0.00
• What Is The Best Practice For Cnns Applied To Visual Instance Retrieval? (2016)0.00
• Visual Similarity Attention (2019)0.00