SegmentAnyTreeV2: Scaling Transformer-Based Tree Instance Segmentation Across Sensors, Platforms, and Forests

• URL: http://arxiv.org/abs/2606.08206v1
• Date: Sat, 06 Jun 2026 14:48:25 GMT
• Title: SegmentAnyTreeV2: Scaling Transformer-Based Tree Instance Segmentation Across Sensors, Platforms, and Forests
• Abstract summary: SegmentAnyTreeV2 is a sensor- and platform-agnostic framework for semantic and instance segmentation of forest point clouds.<n>On the FOR-instanceV2 test split, SegmentAnyTreeV2 achieves 90.5% precision, 80.2% recall, 85.0% F1, 90.7% coverage, and 87.6% semantic mIoU.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: We present SegmentAnyTreeV2, a sensor- and platform-agnostic framework for semantic and instance segmentation of forest point clouds. The model combines a serialization-based Point Transformer v3 backbone with a lightweight semantic head and a tree-focused cross-attention mask decoder. Semantic predictions restrict instance decoding to tree-class voxels, while instance-aware query initialization, one-to-many seed supervision, and asymmetric mask scoring improve separation in dense and structurally complex stands. We further introduce FOR-instance v3, an expanded benchmark comprising 427 scenes and 26,496 annotated trees across diverse biomes, forest structures, and LiDAR platforms. On the FOR-instanceV2 test split, SegmentAnyTreeV2 achieves 90.5% precision, 80.2% recall, 85.0% F1, 90.7% coverage, and 87.6% semantic mIoU, outperforming previous learning-based methods in both instance detection and mask completeness. Zero-shot evaluation on independent sites further demonstrates strong cross-domain generalization.

Related papers

• SelectAnyTree: A Promptable Instance Segmentation Model for 3D Forest LiDAR Point Clouds [12.836432207782039]
  SelectAnyTree is a promptable instance segmentation model that delineates any individual tree in a 3D forest point cloud from a few clicks.<n>It segments a target tree to 78.2 Intersection over Union (IoU) from a single click, 24.8 points above the strongest promptable baseline.<n>It reaches every accuracy target with the fewest clicks, while using far fewer parameters and less inference time than prior promptable models.
  arXiv  Detail & Related papers  (2026-06-25T19:15:36Z)
• INSID3: Training-Free In-Context Segmentation with DINOv3 [63.961377087673476]
  INSID3 is a training-free approach that segments concepts at varying granularities only from frozen DINOv3 features.<n>It achieves state-of-the-art results across one-shot semantic, part, and personalized segmentation, outperforming previous work by +7.5 % mIoU.
  arXiv  Detail & Related papers  (2026-03-30T14:16:37Z)
• ZS-TreeSeg: A Zero-Shot Framework for Tree Crown Instance Segmentation [5.392796525513568]
  Individual tree crown segmentation is an important task in remote sensing for biomass estimation and ecological monitoring.<n>We propose ZSeg, a framework that adapts from two mature tasks.<n>Our framework generalizes robustly across sensor types and canopy.
  arXiv  Detail & Related papers  (2026-01-31T02:48:17Z)
• ForestFormer3D: A Unified Framework for End-to-End Segmentation of Forest LiDAR 3D Point Clouds [0.06282171844772422]
  We present ForestFormer3D, a new unified and end-to-end framework for precise individual tree and semantic segmentation.<n>ForestFormer3D incorporates ISA-guided query point selection, a score-based block merging strategy during inference, and a one-to-many association mechanism for effective training.<n>Our model achieves state-of-the-art performance for individual tree segmentation on the newly introduced FOR-instanceV2 dataset.
  arXiv  Detail & Related papers  (2025-06-20T13:39:27Z)
• Point2Tree(P2T) -- framework for parameter tuning of semantic and instance segmentation used with mobile laser scanning data in coniferous forest [0.22940141855172028]
  We train a model based on Pointnet++ architecture which achieves 0.92 F1-score in semantic segmentation. As a second step in our pipeline we used graph-based approach for instance segmentation which reached F1-score approx. 0.6.
  arXiv  Detail & Related papers  (2023-05-04T08:45:17Z)
• Sparse Instance Activation for Real-Time Instance Segmentation [72.23597664935684]
  We propose a conceptually novel, efficient, and fully convolutional framework for real-time instance segmentation. SparseInst has extremely fast inference speed and achieves 40 FPS and 37.9 AP on the COCO benchmark.
  arXiv  Detail & Related papers  (2022-03-24T03:15:39Z)
• Semantic Attention and Scale Complementary Network for Instance Segmentation in Remote Sensing Images [54.08240004593062]
  We propose an end-to-end multi-category instance segmentation model, which consists of a Semantic Attention (SEA) module and a Scale Complementary Mask Branch (SCMB) SEA module contains a simple fully convolutional semantic segmentation branch with extra supervision to strengthen the activation of interest instances on the feature map. SCMB extends the original single mask branch to trident mask branches and introduces complementary mask supervision at different scales.
  arXiv  Detail & Related papers  (2021-07-25T08:53:59Z)
• Multi-task fully convolutional network for tree species mapping in dense forests using small training hyperspectral data [0.9646880519252493]
  This work proposes a multi-task fully convolutional architecture for tree species mapping in dense forests. Our model implements a partial loss function that enables dense tree semantic labeling outcomes from non-dense training samples. A distance regression complementary task that enforces tree crown boundary constraints substantially improves the model performance.
  arXiv  Detail & Related papers  (2021-06-01T21:10:10Z)
• Unifying Instance and Panoptic Segmentation with Dynamic Rank-1 Convolutions [109.2706837177222]
  DR1Mask is the first panoptic segmentation framework that exploits a shared feature map for both instance and semantic segmentation. As a byproduct, DR1Mask is 10% faster and 1 point in mAP more accurate than previous state-of-the-art instance segmentation network BlendMask.
  arXiv  Detail & Related papers  (2020-11-19T12:42:10Z)
• PointGroup: Dual-Set Point Grouping for 3D Instance Segmentation [111.7241018610573]
  We present PointGroup, a new end-to-end bottom-up architecture for instance segmentation. We design a two-branch network to extract point features and predict semantic labels and offsets, for shifting each point towards its respective instance centroid. A clustering component is followed to utilize both the original and offset-shifted point coordinate sets, taking advantage of their complementary strength. We conduct extensive experiments on two challenging datasets, ScanNet v2 and S3DIS, on which our method achieves the highest performance, 63.6% and 64.0%, compared to 54.9% and 54.4% achieved by former best
  arXiv  Detail & Related papers  (2020-04-03T16:26:37Z)
• PointINS: Point-based Instance Segmentation [117.38579097923052]
  Mask representation in instance segmentation with Point-of-Interest (PoI) features is challenging because learning a high-dimensional mask feature for each instance requires a heavy computing burden. We propose an instance-aware convolution, which decomposes this mask representation learning task into two tractable modules. Along with instance-aware convolution, we propose PointINS, a simple and practical instance segmentation approach.
  arXiv  Detail & Related papers  (2020-03-13T08:24:58Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.