# Learn 3D VQA Better with Active Selection and Reannotation

arXiv: https://arxiv.org/abs/2507.04630 · HTML: https://mlsanity.com/paper/2507.04630/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Placed in Spatial and embodied reasoning, one of 17 directions in LLM reasoning.
- Builds on 12 of its 27 references that are in the map.
- First posted on arXiv 7 Jul 2025.

## Placed in

- LLM reasoning / [Spatial and embodied reasoning](https://mlsanity.com/lr/spatial-and-embodied-reasoning/)

## Builds on (most cited first, 12 of 12)

- [VoxPoser: Composable 3D Value Maps for Robotic Manipulation with Language Models](https://mlsanity.com/paper/2307.05973/) (2023)
- [An Embodied Generalist Agent in 3D World](https://mlsanity.com/paper/2311.12871/) (2023)
- [SQA3D: Situated Question Answering in 3D Scenes](https://mlsanity.com/paper/2210.07474/) (2022)
- [LL3DA: Visual Interactive Instruction Tuning for Omni-3D Understanding, Reasoning, and Planning](https://mlsanity.com/paper/2311.18651/) (2023)
- [3D-VisTA: Pre-trained Transformer for 3D Vision and Text Alignment](https://mlsanity.com/paper/2308.04352/) (2023)
- [Influence Selection for Active Learning](https://mlsanity.com/paper/2108.09331/) (2021)
- [CEREALS - Cost-Effective REgion-based Active Learning for Semantic Segmentation](https://mlsanity.com/paper/1810.09726/) (2018)
- [Exploring Active 3D Object Detection from a Generalization Perspective](https://mlsanity.com/paper/2301.09249/) (2023)
- [Bridging the Gap between 2D and 3D Visual Question Answering: A Fusion Approach for 3D VQA](https://mlsanity.com/paper/2402.15933/) (2024)
- [Active Learning for Deep Visual Tracking](https://mlsanity.com/paper/2110.13259/) (2021)
- [Towards Balanced Active Learning for Multimodal Classification](https://mlsanity.com/paper/2306.08306/) (2023)
- [Ask-n-Learn: Active Learning via Reliable Gradient Representations for Image Classification](https://mlsanity.com/paper/2009.14448/) (2020)

## Built on by (most cited first, 3 of 3)

- [Advancing 3D Scene Understanding with MV-ScanQA Multi-View Reasoning Evaluation and TripAlign Pre-training Dataset](https://mlsanity.com/paper/2508.11058/) (2025)
- [Distilling Neuro-Symbolic Programs into 3D Multi-modal LLMs](https://mlsanity.com/paper/2606.01215/) (2026)
- [Scalable Object Relation Encoding for Better 3D Spatial Reasoning in Large Language Models](https://mlsanity.com/paper/2603.24721/) (2026)

## Abstract (arXiv)

3D Visual Question Answering (3D VQA) is crucial for enabling models to perceive the physical world and perform spatial reasoning. In 3D VQA, the free-form nature of answers often leads to improper annotations that can confuse or mislead models when training on the entire dataset. While other text generation tasks can mitigate this issue by learning on large-scale datasets, the scarcity of 3D scene data enlarges the negative effect of misleading annotations. Although active learning strategies can select valuable instances for training, they fail to identify and resolve misleading labels, which the oracle inevitably provides in practice. To address this issue, we propose a multi-turn interactive active learning strategy. This strategy selects data based on models' semantic uncertainty to form a solid knowledge foundation more effectively and actively requests reannotation from an oracle to resolve potentially misleading labels. For uncertainty assessment, we utilize a variance-based metric that takes semantic relationships between terms into consideration, thus avoiding the uniform inter-class similarity assumption of previous assessment metrics. Extensive experiments exhibit better model performance and a substantial reduction in training costs, with a halving of training costs for achieving relatively high accuracy. The code is available at https://github.com/fz-zsl/AQuA.