# 3rd Place Solution to Human Motion Challenges in Real-World and Clinical Settings (MoCha) @ECCV2026: Language-Aligned Motion Representations for Domain-Generalizable UPDRS-Gait Severity Estimation

arXiv: https://arxiv.org/abs/2609.10187 · HTML: https://mlsanity.com/paper/2609.10187/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 8 of its 9 references that are in the map.
- First posted on arXiv 9 Sep 2026.

## Builds on (most cited first, 8 of 8)

- [Long-tail learning via logit adjustment](https://mlsanity.com/paper/2007.07314/) (2020)
- [Causality matters in medical imaging](https://mlsanity.com/paper/1912.08142/) (2019)
- [CLOC: Contrastive Learning for Ordinal Classification with Multi-Margin N-pair Loss](https://mlsanity.com/paper/2504.17813/) (2025)
- [CARE-PD: A Multi-Site Anonymized Clinical Dataset for Parkinson's Disease Gait Assessment](https://mlsanity.com/paper/2510.04312/) (2025)
- [Bridging Domains through Subspace-Aware Model Merging](https://mlsanity.com/paper/2603.05768/) (2026)
- [Domain Adaptive Diabetic Retinopathy Grading with Model Absence and Flowing Data](https://mlsanity.com/paper/2412.01203/) (2024)
- [Aggregate, Don't Adapt: Subject-Level Posterior Aggregation and Transductive Calibration for Cross-Site Parkinsonian Gait Severity](https://mlsanity.com/paper/2608.20587/) (2026)
- [More Motion Is Not Always Better Motion: Corpus Composition Governs Whether Augmentation Helps SMPL-Based Parkinsonian Gait Severity Estimation](https://mlsanity.com/paper/2608.23730/) (2026)

## Abstract (arXiv)

In this work, we introduce language-aligned motion representations for domain-generalizable UPDRS-Gait severity estimation, aiming to learn semantically structured motion features that generalize across heterogeneous clinical domains. We first learn motion representations using a Bi-GRU backbone that captures the temporal dynamics of SMPL sequences. Prior to model training, motion captions are generated offline using Qwen2.5-7B-Instruct. The backbone is then trained with both classification and text-alignment objectives to learn discriminative and semantically structured motion representations while accounting for the class imbalance present in the training data. We subsequently adapt the learned backbone independently to each source domain so that the model can capture domain-specific motion characteristics. The resulting source-specific models are then merged at the parameter level to consolidate complementary knowledge across source domains into a single domain-generalized model. To further mitigate class imbalance, we perform GPT-5.5-based pseudo labeling, and our final merged models for each site do not use any class-prior correction during inference. The resulting model is evaluated under the unseen-site setting of the MoCha Challenge, using Macro F1 as the primary evaluation metric. Our method achieves a macro-F1 of 0.57 on the hidden test set with only 637K active parameters at inference, ranking 3rd among 58 leaderboard entries in the MoCha 2026 Challenge. The challenge attracted 1,669 submissions from 112 participants and offered monetary prizes sponsored by Machine Medicine Technologies.