Looking for the Devil in the Details: Learning Trilinear Attention Sampling Network for Fine-grained Image Recognition
arXiv:1903.06150
Abstract
Learning subtle yet discriminative features (e.g., beak and eyes for a bird) plays a significant role in fine-grained image recognition. Existing attention-based approaches localize and amplify significant parts to learn fine-grained details, which often suffer from a limited number of parts and heavy computational cost. In this paper, we propose to learn such fine-grained features from hundreds of part proposals by Trilinear Attention Sampling Network (TASN) in an efficient teacher-student manner. Specifically, TASN consists of 1) a trilinear attention module, which generates attention maps by modeling the inter-channel relationships, 2) an attention-based sampler which highlights attended parts with high resolution, and 3) a feature distiller, which distills part features into a global one by weight sharing and feature preserving strategies. Extensive experiments verify that TASN yields the best performance under the same settings with the most competitive approaches, in iNaturalist-2017, CUB-Bird, and Stanford-Cars datasets.
References in corpus (6)
- Distilling the Knowledge in a Neural Network
- MXNet: A Flexible and Efficient Machine Learning Library for Heterogeneous Distributed Systems
- Bird Species Categorization Using Pose Normalized Deep Convolutional Nets
- Fully Convolutional Attention Networks for Fine-Grained Recognition
- Knowledge Distillation with Adversarial Samples Supporting Decision Boundary
- Learning to Zoom: a Saliency-Based Sampling Layer for Neural Networks
Cited by in corpus (5)
- Deep Learning Techniques for In-Crop Weed Identification: A Review
- Tackling the Challenges in Scene Graph Generation with Local-to-Global Interactions
- Human Attention in Fine-grained Classification
- Channel Interaction Networks for Fine-Grained Image Categorization
- Spatial-Scale Aligned Network for Fine-Grained Recognition