# ProGuT: Label-Efficient Panoptic Segmentation for Forest Scenes

arXiv: https://arxiv.org/abs/2609.36891 · HTML: https://mlsanity.com/paper/2609.36891/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 25 of its 39 references that are in the map.
- First posted on arXiv 29 Sep 2026.

## Builds on (most cited first, 25 of 25)

- [Segment Anything](https://mlsanity.com/paper/2304.02643/) (2023)
- [Masked-attention Mask Transformer for Universal Image Segmentation](https://mlsanity.com/paper/2112.01527/) (2021)
- [DINO: DETR with Improved DeNoising Anchor Boxes for End-to-End Object Detection](https://mlsanity.com/paper/2203.03605/) (2022)
- [Per-Pixel Classification is Not All You Need for Semantic Segmentation](https://mlsanity.com/paper/2107.06278/) (2021)
- [Language-driven Semantic Segmentation](https://mlsanity.com/paper/2201.03546/) (2022)
- [Scaling Open-Vocabulary Image Segmentation with Image-Level Labels](https://mlsanity.com/paper/2112.12143/) (2021)
- [Invariant Information Clustering for Unsupervised Image Classification and Segmentation](https://mlsanity.com/paper/1807.06653/) (2018)
- [MaskCLIP: Masked Self-Distillation Advances Contrastive Language-Image Pretraining](https://mlsanity.com/paper/2208.12262/) (2022)
- [Cut and Learn for Unsupervised Object Detection and Instance Segmentation](https://mlsanity.com/paper/2301.11320/) (2023)
- [PiCIE: Unsupervised Semantic Segmentation using Invariance and Equivariance in Clustering](https://mlsanity.com/paper/2103.17070/) (2021)
- [RELLIS-3D Dataset: Data, Benchmarks and Analysis](https://mlsanity.com/paper/2011.12954/) (2020)
- [Unsupervised Semantic Segmentation with Self-supervised Object-centric Representations](https://mlsanity.com/paper/2207.05027/) (2022)
- [The GOOSE Dataset for Perception in Unstructured Environments](https://mlsanity.com/paper/2310.16788/) (2023)
- [Unsupervised Universal Image Segmentation](https://mlsanity.com/paper/2312.17243/) (2023)
- [Unsupervised Semantic Segmentation Through Depth-Guided Feature Correlation and Sampling](https://mlsanity.com/paper/2309.12378/) (2023)
- [WildScenes: A Benchmark for 2D and 3D Semantic Segmentation in Large-scale Natural Environments](https://mlsanity.com/paper/2312.15364/) (2023)
- [CuVLER: Enhanced Unsupervised Object Discoveries through Exhaustive Self-Supervised Transformers](https://mlsanity.com/paper/2403.07700/) (2024)
- [TreeLearn: A deep learning method for segmenting individual trees from ground-based LiDAR forest point clouds](https://mlsanity.com/paper/2309.08471/) (2023)
- [Data or Language Supervision: What Makes CLIP Better than DINO?](https://mlsanity.com/paper/2510.11835/) (2025)
- [Tree Detection and Diameter Estimation Based on Deep Learning](https://mlsanity.com/paper/2210.17424/) (2022)
- [Scene-Centric Unsupervised Panoptic Segmentation](https://mlsanity.com/paper/2504.01955/) (2025)
- [FinnWoodlands Dataset](https://mlsanity.com/paper/2304.00793/) (2023)
- [GO: The Great Outdoors Multimodal Dataset](https://mlsanity.com/paper/2501.19274/) (2025)
- [Detecting and Mapping Trees in Unstructured Environments with a Stereo Camera and Pseudo-Lidar](https://mlsanity.com/paper/2103.15967/) (2021)
- [Beyond Single Stage Encoder-Decoder Networks: Deep Decoders for Semantic Image Segmentation](https://mlsanity.com/paper/2007.09746/) (2020)

## Abstract (arXiv)

Panoptic segmentation in forest environments is bottlenecked not by semantic quality but by instance separation; existing unsupervised panoptic approaches produce usable stuff maps but near-zero thing quality. Depth or flow-based instance discovery methods needs sensors that are not always available. We present ProGuT (Prototype Guided Training), which produces panoptic pseudo-labels without per-image training masks, needing only unlabeled images and one-time cluster-to-class mapping. ProGuT clusters CLIP patch features, then recovers trunk instances through multiscale geometric prior that falsifies non-trunk structures via structure-tensor. This is cheap compared to depth, flow or class-supervision methods to create pseudo labels. These are then used for downstream tasks which we evaluate against other unsupervised baselines. ProGuT achieves a Panoptic Quality (PQ) of 65.2 on Our-forest dataset (2.6x improvement over the initial pseudo-label quality) and reaches 65.9 mIoU on Freiburg Forest, outperforming unsupervised baselines like PiCIE (45.3 IoU) and STEGO(57.6IoU). Additionally, ProGuT outperforms existing unsupervised methods for class-agnostic trunk instance benchmark.