Universal Skeleton Understanding via Differentiable Rendering and MLLMs

• URL: http://arxiv.org/abs/2603.18003v1
• Date: Wed, 18 Mar 2026 17:59:12 GMT
• Title: Universal Skeleton Understanding via Differentiable Rendering and MLLMs
• Abstract summary: SkeletonLLM achieves universal skeleton understanding by translating arbitrary skeleton sequences into the MLLM's native visual modality.<n>Because the pipeline is end-to-end differentiable, MLLM gradients can directly guide the rendering to produce task-informative visual tokens.
• Score: 50.776811193752856
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: Multimodal large language models (MLLMs) exhibit strong visual-language reasoning, yet remain confined to their native modalities and cannot directly process structured, non-visual data such as human skeletons. Existing methods either compress skeleton dynamics into lossy feature vectors for text alignment, or quantize motion into discrete tokens that generalize poorly across heterogeneous skeleton formats. We present SkeletonLLM, which achieves universal skeleton understanding by translating arbitrary skeleton sequences into the MLLM's native visual modality. At its core is DrAction, a differentiable, format-agnostic renderer that converts skeletal kinematics into compact image sequences. Because the pipeline is end-to-end differentiable, MLLM gradients can directly guide the rendering to produce task-informative visual tokens. To further enhance reasoning capabilities, we introduce a cooperative training strategy: Causal Reasoning Distillation transfers structured, step-by-step reasoning from a teacher model, while Discriminative Finetuning sharpens decision boundaries between confusable actions. SkeletonLLM demonstrates strong generalization on diverse tasks including recognition, captioning, reasoning, and cross-format transfer -- suggesting a viable path for applying MLLMs to non-native modalities. Code will be released upon acceptance.

Related papers

• SCALPEL: Semantic Cross-modal Alignment via LLM-Powered Encoder Learning for Medical Vision-Language Representation [5.487826115955642]
  Vision-language pre-training serves as a cornerstone for medical multimodal representation learning.<n>We propose textbfSCALPEL, a textbfSemantic textbfCross-modal textbfAlignment framework via textbfLLM-textbfPowered textbfEncoder textbfLearning.
  arXiv  Detail & Related papers  (2026-07-29T13:12:32Z)
• Learning Trajectory-Aware Multimodal Large Language Models for Video Reasoning Segmentation [55.743701532984126]
  We propose TrajSeg, a simple and unified framework built upon Multimodal Large Language Models.<n>In this work, we introduce bidirectional text-trajectory alignment, where MLLMs accept grounding-intended (text-to-trajectory) and captioning-intended (trajectory-to-text) instructions.<n>This way, MLLMs can benefit from enhanced correspondence and better perceive object trajectories in videos.
  arXiv  Detail & Related papers  (2026-03-23T02:25:51Z)
• Stateful Cross-layer Vision Modulation [19.730096071316876]
  multimodal large language models (MLLMs) widely adopt multi-layer visual feature fusion to enhance visual representation.<n>Existing approaches typically perform static concatenation or weighted aggregation after visual encoding, without intervening in the representation formation process itself.<n>We propose a cross-layer memory-modulated vision framework(SCVM) to address these limitations.
  arXiv  Detail & Related papers  (2026-02-28T13:57:19Z)
• PathFLIP: Fine-grained Language-Image Pretraining for Versatile Computational Pathology [6.821738567680833]
  We propose PathFLIP (Pathology Fine-grained Language-Image Pretraining), a novel framework for holistic Whole Slide Images (WSI) interpretation.<n>PathFLIP decomposes slide-level captions into region-level subcaptions and generates text-conditioned region embeddings to facilitate precise visual-language grounding.
  arXiv  Detail & Related papers  (2025-12-19T14:26:50Z)
• Analyzing Finetuning Representation Shift for Multimodal LLMs Steering [56.710375516257876]
  We propose to map hidden states to interpretable visual and textual concepts.<n>This enables us to more efficiently compare certain semantic dynamics, such as the shift from an original and fine-tuned model.<n>We also demonstrate the use of shift vectors to capture these concepts changes.
  arXiv  Detail & Related papers  (2025-01-06T13:37:13Z)
• An Information Compensation Framework for Zero-Shot Skeleton-based Action Recognition [49.45660055499103]
  Zero-shot human skeleton-based action recognition aims to construct a model that can recognize actions outside the categories seen during training. Previous research has focused on aligning sequences' visual and semantic spatial distributions. We introduce a new loss function sampling method to obtain a tight and robust representation.
  arXiv  Detail & Related papers  (2024-06-02T06:53:01Z)
• Vision-Language Meets the Skeleton: Progressively Distillation with Cross-Modal Knowledge for 3D Action Representation Learning [20.34477942813382]
  Skeleton-based action representation learning aims to interpret and understand human behaviors by encoding the skeleton sequences. We introduce a novel skeleton-based training framework based on Cross-modal Contrastive learning. Our method outperforms the previous methods and achieves state-of-the-art results.
  arXiv  Detail & Related papers  (2024-05-31T03:40:15Z)
• Unified Language-Vision Pretraining in LLM with Dynamic Discrete Visual Tokenization [52.935150075484074]
  We introduce a well-designed visual tokenizer to translate the non-linguistic image into a sequence of discrete tokens like a foreign language. The resulting visual tokens encompass high-level semantics worthy of a word and also support dynamic sequence length varying from the image. This unification empowers LaVIT to serve as an impressive generalist interface to understand and generate multi-modal content simultaneously.
  arXiv  Detail & Related papers  (2023-09-09T03:01:38Z)
• LAC: Latent Action Composition for Skeleton-based Action Segmentation [21.797658771678066]
  Skeleton-based action segmentation requires recognizing composable actions in untrimmed videos. Current approaches decouple this problem by first extracting local visual features from skeleton sequences and then processing them by a temporal model to classify frame-wise actions. We propose Latent Action Composition (LAC), a novel self-supervised framework aiming at learning from synthesized composable motions for skeleton-based action segmentation.
  arXiv  Detail & Related papers  (2023-08-28T11:20:48Z)
• SkeletonMAE: Graph-based Masked Autoencoder for Skeleton Sequence Pre-training [110.55093254677638]
  We propose an efficient skeleton sequence learning framework, named Skeleton Sequence Learning (SSL) In this paper, we build an asymmetric graph-based encoder-decoder pre-training architecture named SkeletonMAE. Our SSL generalizes well across different datasets and outperforms the state-of-the-art self-supervised skeleton-based action recognition methods.
  arXiv  Detail & Related papers  (2023-07-17T13:33:11Z)
• SimMC: Simple Masked Contrastive Learning of Skeleton Representations for Unsupervised Person Re-Identification [63.903237777588316]
  We present a generic Simple Masked Contrastive learning (SimMC) framework to learn effective representations from unlabeled 3D skeletons for person re-ID. Specifically, to fully exploit skeleton features within each skeleton sequence, we first devise a masked prototype contrastive learning (MPC) scheme. Then, we propose the masked intra-sequence contrastive learning (MIC) to capture intra-sequence pattern consistency between subsequences.
  arXiv  Detail & Related papers  (2022-04-21T00:19:38Z)

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.