# PanoEnv: Exploring 3D Spatial Intelligence in Panoramic Environments with Reinforcement Learning

arXiv: https://arxiv.org/abs/2602.21992 · HTML: https://mlsanity.com/paper/2602.21992/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- Builds on 62 of its 70 references that are in the map.
- First posted on arXiv 25 Feb 2026.

## Builds on (most cited first, 30 of 62)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [Qwen2.5-VL Technical Report](https://mlsanity.com/paper/2502.13923/) (2025)
- [Improved Baselines with Visual Instruction Tuning](https://mlsanity.com/paper/2310.03744/) (2023)
- [BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation](https://mlsanity.com/paper/2201.12086/) (2022)
- [Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution](https://mlsanity.com/paper/2409.12191/) (2024)
- [InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuning](https://mlsanity.com/paper/2305.06500/) (2023)
- [InternVL: Scaling up Vision Foundation Models and Aligning for Generic Visual-Linguistic Tasks](https://mlsanity.com/paper/2312.14238/) (2023)
- [VQA: Visual Question Answering](https://mlsanity.com/paper/1505.00468/) (2015)
- [Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling](https://mlsanity.com/paper/2412.05271/) (2024)
- [CLEVR: A Diagnostic Dataset for Compositional Language and Elementary Visual Reasoning](https://mlsanity.com/paper/1612.06890/) (2016)
- [Vision-R1: Incentivizing Reasoning Capability in Multimodal Large Language Models](https://mlsanity.com/paper/2503.06749/) (2025)
- [Thinking in Space: How Multimodal Large Language Models See, Remember, and Recall Spaces](https://mlsanity.com/paper/2412.14171/) (2024)
- [LLaVA-NeXT-Interleave: Tackling Multi-image, Video, and 3D in Large Multimodal Models](https://mlsanity.com/paper/2407.07895/) (2024)
- [DeepSeek-VL2: Mixture-of-Experts Vision-Language Models for Advanced Multimodal Understanding](https://mlsanity.com/paper/2412.10302/) (2024)
- [From Recognition to Cognition: Visual Commonsense Reasoning](https://mlsanity.com/paper/1811.10830/) (2018)
- [TartanAir: A Dataset to Push the Limits of Visual SLAM](https://mlsanity.com/paper/2003.14338/) (2020)
- [Prometheus 2: An Open Source Language Model Specialized in Evaluating Other Language Models](https://mlsanity.com/paper/2405.01535/) (2024)
- [Curriculum Learning for Reinforcement Learning Domains: A Framework and Survey](https://mlsanity.com/paper/2003.04960/) (2020)
- [What's "up" with vision-language models? Investigating their struggle with spatial reasoning](https://mlsanity.com/paper/2310.19785/) (2023)
- [How Abilities in Large Language Models are Affected by Supervised Fine-tuning Data Composition](https://mlsanity.com/paper/2310.05492/) (2023)
- [Teacher-Student Curriculum Learning](https://mlsanity.com/paper/1707.00183/) (2017)
- [Reverse Curriculum Generation for Reinforcement Learning](https://mlsanity.com/paper/1707.05300/) (2017)
- [Visual Question Answering: A Survey of Methods and Datasets](https://mlsanity.com/paper/1607.05910/) (2016)
- [ViewSpatial-Bench: Evaluating Multi-perspective Spatial Localization in Vision-Language Models](https://mlsanity.com/paper/2505.21500/) (2025)

## Built on by (most cited first, 7 of 7)

- [PanoWorld: Towards Spatial Supersensing in 360^circ Panorama World](https://mlsanity.com/paper/2605.13169/) (2026)
- [Post-Training in End-to-End Autonomous Driving](https://mlsanity.com/paper/2607.08072/) (2026)
- [Panoramic Scene Understanding: A Survey from Distortion-Aware Engineering to Sphere-Native Modeling](https://mlsanity.com/paper/2606.27745/) (2026)
- [EAGLE-360: Embodied Active Global-to-Local Exploration in 360^circ](https://mlsanity.com/paper/2607.02479/) (2026)
- [OneCanvas: 3D Scene Understanding via Panoramic Reprojection](https://mlsanity.com/paper/2606.19253/) (2026)
- [OmniCoT: A Benchmark for Global and Multi-Step Panoramic Reasoning](https://mlsanity.com/paper/2606.30378/) (2026)
- [SphMind: Towards Robust, Training-Free VLM-based Spatial Reasoning with a 360 Camera](https://mlsanity.com/paper/2609.33462/) (2026)

## Abstract (arXiv)

360 panoramic images are increasingly used in virtual reality, autonomous driving, and robotics for holistic scene understanding. However, current Vision-Language Models (VLMs) struggle with 3D spatial reasoning on Equirectangular Projection (ERP) images due to geometric distortion and limited 3D supervision. We introduce PanoEnv, a large-scale VQA benchmark built from synthetic 3D environments, containing 14.8K questions across five categories (e.g., relative position, volume comparison) grounded in accurate 3D annotations including depth, segmentation, and bounding boxes. Benchmarking 14 state-of-the-art VLMs reveals limited 3D understanding, achieving only 49.34% overall accuracy and 8.36% on open-ended (OE) questions. To enhance 3D reasoning, we propose a reinforcement learning post-training framework based on Group Relative Policy Optimization (GRPO) with a ground-truth-guided reward that incorporates five geometry-aware strategies such as distance tolerance and spatial consistency. A two-stage curriculum further mitigates catastrophic forgetting: Stage 1 trains on structured tasks (true/false and multiple choice), and Stage 2 fine-tunes on mixed open-ended data to improve generalization. Our 7B model achieves new state-of-the-art performance, improving overall accuracy to 52.93% (+3.59%) and open-ended accuracy to 14.83% while maintaining structured-task performance. It also achieves top semantic evaluation scores (Q-Score 6.24, P-Score 5.95), surpassing 32B models. These results demonstrate that PanoEnv-QA and our curriculum-based RL framework effectively instill 3D spatial intelligence in VLMs for omnidirectional perception.