MAETrack: Unleashing the Potential of Pretrained Geometric Priors for 3D Single Object Tracking

• URL: http://arxiv.org/abs/2609.16695v1
• Date: Tue, 15 Sep 2026 06:21:44 GMT
• Title: MAETrack: Unleashing the Potential of Pretrained Geometric Priors for 3D Single Object Tracking
• Abstract summary: Large-scale pre-training has transformed representation learning in 2D vision, yet its transferability to 3D single object tracking (SOT) remains insufficiently understood.<n>We propose MAETrack, a lightweight adaptation framework for transferring pre-training MAE representations to 3D SOT.
• Score: 37.383672908246965
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: Large-scale pre-training has transformed representation learning in 2D vision, yet its transferability to 3D single object tracking (SOT) remains insufficiently understood. Directly fine-tuning self-supervised 3D encoders, such as masked autoencoders (MAE), often leads to sub-optimal adaptation because the reconstruction objective is not fully aligned with the spatial-temporal matching requirements of tracking. In this paper, we observe that this difficulty can be interpreted as a layer-wise transfer mismatch: shallow layers tend to preserve transferable geometric cues, while deeper layers become increasingly specialized to the reconstruction pretext task and are less suitable for downstream tracking. Based on this observation, we propose MAETrack, a lightweight adaptation framework for transferring pre-training MAE representations to 3D SOT. MAETrack includes Layer-Selective Initialization (LSI), which initializes only the shallow stages of the tracking backbone from pre-trained weights while re-initializing deeper stages, and Geometric Residual Gating (GRG), which reinforces structurally salient regions in the search BEV features before template-search fusion through residual spatial modulation. Extensive experiments on standard 3D SOT benchmarks show that MAETrack consistently improves upon vanilla fine-tuning baselines with limited computational overhead. More broadly, our results suggest that effective transfer from 3D reconstruction pre-training to 3D tracking is not merely a matter of partial fine-tuning, but depends on a tracking-oriented transfer principle that preserves shallow geometry while adapting deeper representations to the downstream objective.

Related papers

• Lift3D-VLA: Lifting VLA Models to 3D Geometry and Dynamics-Aware Manipulation [76.16065370700488]
  We introduce Lift3D-VLA, a unified VLA framework that equips models with explicit 3D point cloud reasoning and enables temporally coherent action generation.<n>We show that Lift3D-VLA achieves 10.8% and 11.1% higher mean success rates on MetaWorld and RLBench than the best-performing prior VLA methods.
  arXiv  Detail & Related papers  (2026-07-07T17:59:47Z)
• GAP-MLLM: Geometry-Aligned Pre-training for Activating 3D Spatial Perception in Multimodal Large Language Models [70.61152292499737]
  We argue that this gap does not arise from insufficient geometric priors, but from a misalignment in the training paradigm.<n>Existing approaches typically resort to naive feature concatenation and optimize directly for downstream tasks without geometry-specific supervision.<n>We propose GAP-MLLM, a Geometry-Aligned Pre-training paradigm that explicitly activates structural perception before downstream adaptation.
  arXiv  Detail & Related papers  (2026-03-17T12:43:48Z)
• Tail-Aware Post-Training Quantization for 3D Geometry Models [58.79500829118265]
  Post-Training Quantization (PTQ) enables efficient inference without retraining.<n>PTQ fails to transfer effectively to 3D models due to intricate feature distributions and prohibitive calibration overhead.<n>We propose TAPTQ, a Tail-Aware Post-Training Quantization pipeline for 3D geometric learning.
  arXiv  Detail & Related papers  (2026-02-02T07:21:15Z)
• TrackAny3D: Transferring Pretrained 3D Models for Category-unified 3D Point Cloud Tracking [25.788917457593673]
  TrackAny3D is the first framework to transfer large-scale pretrained 3D models for category-agnostic 3D SOT.<n>MoGE architecture adaptively activates specialized 3works based on distinct geometric characteristics.<n>Experiments show that TrackAny3D establishes new state-of-the-art performance on category-agnostic 3D SOT.
  arXiv  Detail & Related papers  (2025-07-26T10:41:55Z)
• T-3DGS: Removing Transient Objects for 3D Scene Reconstruction [83.05271859398779]
  Transient objects in video sequences can significantly degrade the quality of 3D scene reconstructions.<n>We propose T-3DGS, a novel framework that robustly filters out transient distractors during 3D reconstruction using Gaussian Splatting.
  arXiv  Detail & Related papers  (2024-11-29T07:45:24Z)
• S.T.A.R.-Track: Latent Motion Models for End-to-End 3D Object Tracking with Adaptive Spatio-Temporal Appearance Representations [10.46571824050325]
  Following the tracking-by-attention paradigm, this paper introduces an object-centric, transformer-based framework for tracking in 3D. Inspired by this, we propose S.T.A.R.-Track, which uses a novel latent motion model (LMM) to adjust object queries to account for changes in viewing direction and lighting conditions directly in the latent space.
  arXiv  Detail & Related papers  (2023-06-30T12:22:41Z)
• Geometry-Contrastive Transformer for Generalized 3D Pose Transfer [95.56457218144983]
  The intuition of this work is to perceive the geometric inconsistency between the given meshes with the powerful self-attention mechanism. We propose a novel geometry-contrastive Transformer that has an efficient 3D structured perceiving ability to the global geometric inconsistencies. We present a latent isometric regularization module together with a novel semi-synthesized dataset for the cross-dataset 3D pose transfer task.
  arXiv  Detail & Related papers  (2021-12-14T13:14:24Z)
• Learning monocular 3D reconstruction of articulated categories from motion [39.811816510186475]
  Video self-supervision forces the consistency of consecutive 3D reconstructions by a motion-based cycle loss. We introduce an interpretable model of 3D template deformations that controls a 3D surface through the displacement of a small number of local, learnable handles. We obtain state-of-the-art reconstructions with diverse shapes, viewpoints and textures for multiple articulated object categories.
  arXiv  Detail & Related papers  (2021-03-30T13:50:27Z)

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.