Research map / Multi-agent systems and games / LLM agents
NutriBench-Kitchen: Benchmarking Embodied AI for Nutrition Management
YuLin Wei, Xiangchen Wang, Jianhui Pan, Jinyu Xiao, Zheng Tan, Ruozai Tian, Guanhua Chen, Feng Zheng
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add Nutri-Vgent structured memory | Qwen3-VL-8B · NutriBench-Kitchen | average accuracy (%): 53.4 -> 62.3 | Improved |
| add Nutri-Vgent structured memory | Qwen3-VL-4B · NutriBench-Kitchen | average accuracy (%): 55.5 -> 62.1 | Improved |
| add generic Vgent RAG | Qwen3-VL-8B · NutriBench-Kitchen | average accuracy (%): 53.4 -> 31.7 | Hurt |
| add generic Vgent RAG | Qwen3-VL-4B · NutriBench-Kitchen | average accuracy (%): 55.5 -> 28.8 | Hurt |
| add Nutri-Vgent memory | Qwen3-VL-8B · NutriBench-Kitchen RQ Hard | accuracy (%): 63.4 -> 94.1 | Improved |
| add Nutri-Vgent memory | Qwen3-VL-8B · NutriBench-Kitchen IE Hard | accuracy (%): 14.0 -> 73.3 | Improved |
| add Nutri-Vgent memory | Qwen3-VL-8B · NutriBench-Kitchen IE Easy | accuracy (%): 30.7 -> 20.0 | Hurt |
| add Episodic Memory only to visual input (ablation) | Qwen3-VL-8B · NutriBench-Kitchen (memory ablation) | IE (%): 22.35 -> 57.00 | Improved |
| add Food Memory only to visual input (ablation) | Qwen3-VL-8B · NutriBench-Kitchen (memory ablation) | overall (%): 53.4 -> 46.9 | Hurt |
| add Episodic Memory only to visual input (ablation) | Qwen3-VL-8B · NutriBench-Kitchen (memory ablation) | overall (%): 53.4 -> 51.8 | Hurt |
| add both Episodic and Food Memory (ablation) | Qwen3-VL-8B · NutriBench-Kitchen (memory ablation) | RQ (%): 67.05 -> 76.05 | Improved |
| add both Episodic and Food Memory (ablation) | Qwen3-VL-8B · NutriBench-Kitchen (memory ablation) | STP (%): 53.00 -> 72.30 | Improved |
Closest papers by meaning
- Pic2Prep: A Multimodal Conversational Agent for Cooking Assistance (AAAI 2025 · similarity 0.88)
- EatVid-Bench: A Multimodal Fine-Grained Eating Behavior Video Dataset (ECCV 2026 · similarity 0.86)
- EmbodiedBench: Comprehensive Benchmarking Multi-modal Large Language Models for Vision-Driven Embodied Agents (ICML 2025 · similarity 0.84)
- TaskGround: Structured Executable Task Inference for Full-Scene Household Reasoning (NeurIPS 2026 · similarity 0.83)
- Embodied Web Agents: Bridging Physical-Digital Realms for Integrated Agent Intelligence (NeurIPS 2025 · similarity 0.83)
- NutriBench: A Dataset for Evaluating Large Language Models in Nutrition Estimation from Meal Descriptions (ICLR 2025 · similarity 0.82)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
