The Unreasonable Effectiveness of Noisy Data for Fine-Grained Recognition

Abstract

Current approaches for fine-grained recognition do the following: First, recruit experts to annotate a dataset of images, optionally also collecting more structured data in the form of part annotations and bounding boxes. Second, train a model utilizing this data. Toward the goal of solving fine-grained recognition, we introduce an alternative approach, leveraging free, noisy data from the web and simple, generic methods of recognition. This approach has benefits in both performance and scalability. We demonstrate its efficacy on four fine-grained datasets, greatly exceeding existing state of the art without the manual collection of even a single label, and furthermore show first results at scaling to more than 10,000 fine-grained categories. Quantitatively, we achieve top-1 accuracies of 92.3% on CUB-200-2011, 85.4% on Birdsnap, 93.4% on FGVC-Aircraft, and 80.8% on Stanford Dogs without using their annotated training sets. We compare our approach to an active learning approach for expanding fine-grained datasets.

ECCV 2016, data is released

References in corpus (7)

• Training Deep Neural Networks on Noisy Labels with Bootstrapping
• Training Convolutional Networks with Noisy Labels
• Bird Species Categorization Using Pose Normalized Deep Convolutional Nets
• Attention for Fine-Grained Categorization
• Part Detector Discovery in Deep Convolutional Neural Networks
• Importance Weighted Active Learning
• Hierarchical Subquery Evaluation for Active Learning on a Graph

Cited by in corpus (15)

• MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications
• Revisiting Unreasonable Effectiveness of Data in Deep Learning Era
• Learning from Noisy Labels with Distillation
• Deep Convolutional Neural Network Design Patterns
• Active and Continuous Exploration with Deep Neural Networks and Expected Model Output Changes
• Rethinking generalization requires revisiting old ideas: statistical mechanics approaches and complex learning behavior
• Fine-Grained Car Detection for Visual Census Estimation
• Borrowing Treasures from the Wealthy: Deep Transfer Learning through Selective Joint Fine-tuning
• Joint Multi-Person Pose Estimation and Semantic Part Segmentation
• Fine-grained Recognition in the Wild: A Multi-Task Domain Adaptation Approach
• Attend in groups: a weakly-supervised deep learning framework for learning from web data
• Visual Concept Recognition and Localization via Iterative Introspection
• Low-rank Bilinear Pooling for Fine-Grained Classification
• Automatic Dataset Augmentation
• Scalable Annotation of Fine-Grained Categories Without Experts