FoodSense - Texture: leaderboard
FoodSense (2026) cross-sensory inference for texture: from a food photo the model predicts a 1-5 rating, compared with the mean human rating on 438 test images by Lin's concordance correlation coefficient (-1 to 1, shown times 100); higher is better.
Metric: Texture: Lin's concordance correlation coefficient between the model's predicted 1-5 rating and the mean human rating (about 22 annotators per image) on FoodSense's 438 held-out test food images, zero-shot with the authors' sensory-assessment prompt, ratings parsed by regex; range -1 to 1, shown times 100; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 5 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemma 3 27B (IT) | 18.1 | #1324 |
| 2 | LLaVA-v1.6-34B | 11.3 | #5950 |
| 3 | food-Llama-3.2-11B-Vision-Instruct | 9.7 | |
| 4 | InternVL2.5-26B | 8.6 | #2156 |
| 5 | Qwen 2.5 VL 32B Instruct | 7.8 | #1062 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=foodsense-texture · How It Works · Data refreshed daily, snapshot 2026-11-21.