Reproduce “TactileEval: A Step Towards Automated Fine-Grained Evaluation and Editing of Tactile Graphics” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Adnan Khan, Abbas Akkasi, Majid Komeili
TactileEval turns expert feedback into five defect categories and uses a feature probe to guide edits to existing tactile graphics. The probe encodes paired natural and tactile images with CLIP ViT-L/14, reaching 85.70% accuracy on held-out option-level records. In selected edits, its issue score fell in 14 of 15 cases; this same-probe result is not user validation.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.