# Can Frozen Hyperspherical Features Guide the Selection of Pseudo Masks?

arXiv: https://arxiv.org/abs/2609.30080 · HTML: https://mlsanity.com/paper/2609.30080/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 46 of its 69 references that are in the map.
- First posted on arXiv 24 Sep 2026.

## Builds on (most cited first, 30 of 46)

- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [Segment Anything](https://mlsanity.com/paper/2304.02643/) (2023)
- [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023)
- [Emerging Properties in Self-Supervised Vision Transformers](https://mlsanity.com/paper/2104.14294/) (2021)
- [Qwen2.5-VL Technical Report](https://mlsanity.com/paper/2502.13923/) (2025)
- [On Calibration of Modern Neural Networks](https://mlsanity.com/paper/1706.04599/) (2017)
- [FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence](https://mlsanity.com/paper/2001.07685/) (2020)
- [Vision Transformers for Dense Prediction](https://mlsanity.com/paper/2103.13413/) (2021)
- [MixMatch: A Holistic Approach to Semi-Supervised Learning](https://mlsanity.com/paper/1905.02249/) (2019)
- [SAM 3: Segment Anything with Concepts](https://mlsanity.com/paper/2511.16719/) (2025)
- [Structure-measure: A New Way to Evaluate Foreground Maps](https://mlsanity.com/paper/1708.00786/) (2017)
- [Enhanced-alignment Measure for Binary Foreground Map Evaluation](https://mlsanity.com/paper/1805.10421/) (2018)
- [Mask Scoring R-CNN](https://mlsanity.com/paper/1903.00241/) (2019)
- [Localizing Objects with Self-Supervised Transformers and no Labels](https://mlsanity.com/paper/2109.14279/) (2021)
- [Anabranch Network for Camouflaged Object Segmentation](https://mlsanity.com/paper/2105.09451/) (2021)
- [The Secrets of Salient Object Segmentation](https://mlsanity.com/paper/1406.2807/) (2014)
- [Simultaneously Localize, Segment and Rank the Camouflaged Objects](https://mlsanity.com/paper/2103.04011/) (2021)
- [Cut and Learn for Unsupervised Object Detection and Instance Segmentation](https://mlsanity.com/paper/2301.11320/) (2023)
- [Visual Saliency Based on Multiscale Deep Features](https://mlsanity.com/paper/1503.08663/) (2015)
- [Self-Supervised Transformers for Unsupervised Object Discovery using Normalized Cut](https://mlsanity.com/paper/2202.11539/) (2022)
- [Highly Accurate Dichotomous Image Segmentation](https://mlsanity.com/paper/2203.03041/) (2022)
- [Progressive Identification of True Labels for Partial-Label Learning](https://mlsanity.com/paper/2002.08053/) (2020)
- [Salient Object Detection: A Discriminative Regional Feature Integration Approach](https://mlsanity.com/paper/1410.5926/) (2014)
- [Reti-Diff: Illumination Degradation Image Restoration with Retinex-based Latent Diffusion Model](https://mlsanity.com/paper/2311.11638/) (2023)
- [Unsupervised Object Localization: Observing the Background to Discover Objects](https://mlsanity.com/paper/2212.07834/) (2022)
- [Pyramid Grafting Network for One-Stage High Resolution Saliency Detection](https://mlsanity.com/paper/2204.05041/) (2022)
- [Unsupervised Salient Object Detection with Spectral Cluster Voting](https://mlsanity.com/paper/2203.12614/) (2022)
- [A Survey of Camouflaged Object Detection and Beyond](https://mlsanity.com/paper/2408.14562/) (2024)
- [Reverse Classification Accuracy: Predicting Segmentation Performance in the Absence of Ground Truth](https://mlsanity.com/paper/1702.03407/) (2017)
- [Learning Noise-Aware Encoder-Decoder from Noisy Labels by Alternating Back-Propagation for Saliency Detection](https://mlsanity.com/paper/2007.12211/) (2020)

## Abstract (arXiv)

Foundation segmenters such as SAM return several plausible masks for an unlabeled image, and a student trained on the wrong one inherits its errors. Choosing among them means querying a second large model or fitting a quality head to annotated masks. We show that a candidate can be judged by what it does to a frozen self-supervised backbone's features. Normalized DINOv2 patch features lie on a hypersphere, and a candidate mask splits that sphere in two. Based on this reading, we introduce SphereTrust, which scores each candidate by three properties of the split, the angular contrast between the two sides, the coverage of the foreground's appearance modes, and contact with the image frame, one for each of three common ways a mask fails, and ranks a pool in 0.55 s per image from the frozen features alone. On eight SAM and SAM3 candidate pools spanning camouflaged, salient, and dichotomous segmentation and camouflage under low light, SphereTrust exceeds the strongest evaluated external baseline on six pools by 1.7 to 9.3 percentage points in mean selected Dice. These comparisons include published selection rules and explicitly labeled adaptations of DSS and UCOD-MKD. On the two prompted camouflage pools, its mean selected Dice is within 0.1 percentage points of the candidate-derived DSS adaptation, with a lower catastrophic-error rate. Which cue carries the signal depends on the candidate pool. The same sphere also supports training. The leading candidates enter as a candidate set with their scores as priors, prototypes reorder them, and a cross-fitted second round completes the labels, raising weighted F by 4.5, 2.3, and 5.5 points over fixed-label training on the three MLLM anchor pools, with students competitive with published unsupervised methods on nineteen test sets.