What We Miss Matters: Learning from the Overlooked
in Point Cloud Transformers

Abstract

Point Cloud Transformers have become a cornerstone in 3D representation fortheir ability to model long-range dependencies via self-attention. However, these models tend to overemphasize salient regions while neglecting other informative regions, which limits feature diversity and compromises robustness. To address this challenge, we introduce BlindFormer, a novel contrastive attention learning framework that redefines saliency by explicitly incorporating features typically neglected by the model. The proposed Attentional Blindspot Mining (ABM) suppresses highly attended regions during training, thereby guiding the model to explore its own blind spots. This redirection of attention expands the model’s perceptual field and uncovers richer geometric cues. To consolidate these overlooked features, BlindFormer employs Blindspot-Aware Joint Optimization (BJO), a joint learning objective that integrates blindspot feature alignment with the original pretext task. BJO enhances feature discrimination while preserving performance on the primary task, leading to more robust and generalizable representations.

We validate BlindFormer on several challenging benchmarks and demonstrate consistent performance gains across multiple Transformer backbones. Notably, it improves Point-MAE by +13.4% and PointGPT-S by +6.3% on OBJBG under Gaussian noise. These results highlight the importance of mitigating attentional biases in 3D representation learning, revealing BlindFormer’s superior ability to handle perturbations and improve feature discrimination.

BibTeX

@article{wang2026we,
  title={What We Miss Matters: Learning from the Overlooked in Point Cloud Transformers},
  author={Wang, Yi and Wang, Jiaze and Guo, Ziyu and Zhang, Renrui and Zhou, Donghao and Chen, Guangyong and Liu, Anfeng and Heng, Pheng-Ann},
  journal={Advances in Neural Information Processing Systems},
  volume={38},
  pages={2550--2574},
  year={2026}
}