# Thinking Outside the Pool: Active Training Image Creation for Relative Attributes

arXiv: https://arxiv.org/abs/1901.02551 · HTML: https://mlsanity.com/paper/1901.02551/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 8 Jan 2019.

## Built on by (most cited first, 5 of 5)

- [Probabilistic Compositional Embeddings for Multimodal Image Retrieval](https://mlsanity.com/paper/2204.05845/) (2022)
- [Image Search with Text Feedback by Additive Attention Compositional Learning](https://mlsanity.com/paper/2203.03809/) (2022)
- [Field-Guide-Inspired Zero-Shot Learning](https://mlsanity.com/paper/2108.10967/) (2021)
- [Shapes as Product Differentiation: Neural Network Embedding in the Analysis of Markets for Fonts](https://mlsanity.com/paper/2107.02739/) (2021)
- [Click-Feedback Retrieval](https://mlsanity.com/paper/2305.00052/) (2023)

## Abstract (arXiv)

Current wisdom suggests more labeled image data is always better, and obtaining labels is the bottleneck. Yet curating a pool of sufficiently diverse and informative images is itself a challenge. In particular, training image curation is problematic for fine-grained attributes, where the subtle visual differences of interest may be rare within traditional image sources. We propose an active image generation approach to address this issue. The main idea is to jointly learn the attribute ranking task while also learning to generate novel realistic image samples that will benefit that task. We introduce an end-to-end framework that dynamically "imagines" image pairs that would confuse the current model, presents them to human annotators for labeling, then improves the predictive model with the new examples. With results on two datasets, we show that by thinking outside the pool of real images, our approach gains generalization accuracy for challenging fine-grained attribute comparisons.