Rethinking the Pointer Loss in Table Structure Recognition: Geometry-Aware Pointer Loss for Spatial Locality

• URL: http://arxiv.org/abs/2606.18721v1
• Date: Wed, 17 Jun 2026 05:59:31 GMT
• Title: Rethinking the Pointer Loss in Table Structure Recognition: Geometry-Aware Pointer Loss for Spatial Locality
• Abstract summary: Table Structure Recognition using a pointer network achieves impressive results by predicting HTML sequences while aligning tags to detected text (or cell) regions.<n>When pointer networks fail, 79.6% of errors occur between spatially adjacent cells.<n>We propose Geometry-Aware Pointer (GAP) Loss, which reweights the cross-entropy objective based on spatial proximity to ground truth.
• Score: 12.986158862036866
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Table Structure Recognition (TSR) using a pointer network achieves impressive results by predicting HTML sequences while aligning tags to detected text (or cell) regions. However, our analysis reveals that when pointer networks fail, 79.6% of errors occur between spatially adjacent cells (Manhattan distance <= 2). Despite this, standard cross-entropy loss weights all negative candidates equally. In this work, we propose Geometry-Aware Pointer (GAP) Loss, which reweights the cross-entropy objective based on spatial proximity to ground truth. By applying inverse distance weighting, GAP focuses gradient flow where the model struggles most: immediate neighbors receive stronger gradients than distant cells. Our approach requires only a straightforward modification to the loss computation, maintaining the same model architecture with zero additional inference cost. Extensive experiments on PubTabNet and SynthTabNet demonstrate that GAP consistently reduces adjacent-cell errors, achieving new state-of-the-art performance. Our findings suggest that incorporating geometric inductive biases at the loss level provides a simple yet effective approach to robust TSR. Our code is available at https://github.com/teamreboott/GAP

Related papers

• SimFuse3D: Source-Guided Target Simulation and Confidence-Guided Multi-Stage Localization Reweighting for Cross-Platform 3D Object Detection [14.526062661334437]
  Changes in sensor height and viewpoint alter object-level point distributions, making cross-platform LiDAR unsupervised domain adaptation difficult.<n>We introduce SimFuse3D, which preserves the target placement and repairs the associated pseudo-object using measured geometry from labeled source scans.
  arXiv  Detail & Related papers  (2026-09-04T08:41:56Z)
• LCPNet: Latent Consistent Proximal Unfolding Network for Infrared Small Target Detection [50.3892176822492]
  Infrared small target detection (IRSTD) aims to identify long distance small targets from complex infrared backgrounds.<n>Deep learning methods have improved IRSTD by learning discriminative image-to-mask mappings.<n>Deep unfolding methods partially address this issue by embedding model-driven iterations into neural networks.<n>We propose Latent Consistent Proximal unfolding network (LCPNet) to address these limitations.
  arXiv  Detail & Related papers  (2026-07-06T02:15:09Z)
• Angle-I2P: Angle-Consistent-Aware Hierarchical Attention for Cross-Modality Outlier Rejection [20.302748104301525]
  Image-to-point-cloud registration (I2P) is a fundamental task in robotic applications such as manipulation,grasping, and localization.<n>We propose Angle-I2P, an outlier rejection network that leverages angle-consistent geometric constraints and hierarchical attention.<n>We show that our method achieves state-of-the-art performance on the 7Scenes, RGBD Scenes V2, and a self-collected dataset.
  arXiv  Detail & Related papers  (2026-05-06T06:41:56Z)
• Robust Scene Coordinate Regression via Geometrically-Consistent Global Descriptors [52.57327385675752]
  We propose an aggregator module that learns global descriptors consistent with both geometrical structure and visual similarity.<n>This corrects erroneous associations caused by unreliable overlap scores.<n>Experiments on challenging benchmarks show substantial localization gains in large-scale environments.
  arXiv  Detail & Related papers  (2025-12-19T04:24:03Z)
• AEGIS: Authentic Edge Growth In Sparsity for Link Prediction in Edge-Sparse Bipartite Knowledge Graphs [0.8594140167290097]
  AEGIS is an edge-only augmentation framework that resamples existing training edges.<n>On text-rich GDP graph, semantic KNN achieves the largest AUC improvement and Brier score reduction.
  arXiv  Detail & Related papers  (2025-09-26T07:51:40Z)
• Steerable Pyramid Weighted Loss: Multi-Scale Adaptive Weighting for Semantic Segmentation [0.4662017507844857]
  We propose a novel steerable pyramid-based weighted (SPW) loss function that efficiently generates adaptive weight maps.<n>Our results demonstrate that the proposed SPW loss function achieves superior pixel precision and segmentation accuracy with minimal computational overhead.
  arXiv  Detail & Related papers  (2025-03-09T13:15:01Z)
• REGRACE: A Robust and Efficient Graph-based Re-localization Algorithm using Consistency Evaluation [23.41000678070751]
  Loop closures are essential for correcting odometry drift and creating consistent maps.<n>Current methods using dense point clouds for accurate place recognition do not scale well due to computationally expensive scan-to-scan comparisons.<n>We introduce REGRACE, a novel approach that addresses these challenges of scalability and perspective difference in re-localization.
  arXiv  Detail & Related papers  (2025-03-05T15:32:38Z)
• iSDF: Real-Time Neural Signed Distance Fields for Robot Perception [64.80458128766254]
  iSDF is a continuous learning system for real-time signed distance field reconstruction. It produces more accurate reconstructions and better approximations of collision costs and gradients.
  arXiv  Detail & Related papers  (2022-04-05T15:48:39Z)
• Scaling Structured Inference with Randomization [64.18063627155128]
  We propose a family of dynamic programming (RDP) randomized for scaling structured models to tens of thousands of latent states. Our method is widely applicable to classical DP-based inference. It is also compatible with automatic differentiation so can be integrated with neural networks seamlessly.
  arXiv  Detail & Related papers  (2021-12-07T11:26:41Z)
• Dense Label Encoding for Boundary Discontinuity Free Rotation Detection [69.75559390700887]
  This paper explores a relatively less-studied methodology based on classification. We propose new techniques to push its frontier in two aspects. Experiments and visual analysis on large-scale public datasets for aerial images show the effectiveness of our approach.
  arXiv  Detail & Related papers  (2020-11-19T05:42:02Z)
• Making Affine Correspondences Work in Camera Geometry Computation [62.7633180470428]
  Local features provide region-to-region rather than point-to-point correspondences. We propose guidelines for effective use of region-to-region matches in the course of a full model estimation pipeline. Experiments show that affine solvers can achieve accuracy comparable to point-based solvers at faster run-times.
  arXiv  Detail & Related papers  (2020-07-20T12:07:48Z)
• Pixel-in-Pixel Net: Towards Efficient Facial Landmark Detection in the Wild [104.61677518999976]
  We propose Pixel-in-Pixel Net (PIPNet) for facial landmark detection. The proposed model is equipped with a novel detection head based on heatmap regression. To further improve the cross-domain generalization capability of PIPNet, we propose self-training with curriculum.
  arXiv  Detail & Related papers  (2020-03-08T12:23:42Z)

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.