DTD

The DTD (Describable Textures Dataset) is a benchmark that contains a collection of texture images used to evaluate the performance of vision-language models in recognizing and generalizing across diverse texture categories.

Papers using DTD (2)

Replication Study: Federated Text-Driven Prompt Generation for Vision-Language Models2026Leveraging Foundation Models for Efficient Federated Learning in Resource-restricted Edge Networks2024 · 1 cites