arXiv Machine Learning August 7, 2026 By Lala Shakti Swarup Ray, Vitor Fortes Rey, Mengxi Liu, Paul Lukowicz, Bo Zhou

VSMP-IMU: Video-Grounded Semantic Motion Programs for Sensor-Aware Synthetic IMU Generation

Read the original on arXiv Machine Learning →

arXiv:2608. 05782v1 Announce Type: cross Abstract: Wearable human activity recognition (HAR) is often limited by the scarcity of labeled sensor data, especially in low-resource, class-imbalanced, and subject-generalization settings.

Summary generated by The Flow from the publisher's feed. The full article lives at arXiv Machine Learning.

benchmarks

arXiv AI

Jun 4

Gravity-Aware Hierarchical Routing for Lightweight SensorLLM on Human Activity Recognition

arXiv:2606. 04019v1 Announce Type: cross Abstract: Recent studies on sensor-language alignment have shown that two-stage frameworks can improve the semantic modeling ability of wearable-sensor human activity recognition (HAR), where SensorLLM-style methods first perform motion-to-language alignment and then fine-tune the model for downstream tasks.

By Hao Li, Mingrui Zheng, Yasuyuki Tahara, Yuichi Sei

nlp fine-tuning safety

arXiv Machine Learning

Jul 2

ActivityNarrated: An Open-Ended Narrative Paradigm for Wearable Human Activity Understanding

arXiv:2604. 00767v2 Announce Type: replace Abstract: Wearable human activity recognition (HAR) has made steady progress, yet much of this progress remains grounded in fixed-window, closed-set classification benchmarks.

By Lala Shakti Swarup Ray, Mengxi Liu, Alcina Pinto, Deepika Gurung, Daniel Geissler, Paul Lukowoicz, Bo Zhou

llms benchmarks safety

arXiv Machine Learning

1d ago

TransfHAR: Self-Supervised Wrist Representations for On-Demand Activity Recognition

arXiv:2608. 15861v1 Announce Type: new Abstract: Fine-grained wrist activity recognition can support applications such as procedural step guidance and context-aware assistance, yet acquiring labeled data for every new task, user, and activity granularity remains a bottleneck.

By Aidan Bradshaw, Riku Arakawa, Xin Liu, Karan Ahuja

arXiv AI

Jun 4

Can VLMs Predict Future States? Bootstrapping World Models from Inverse Dynamics

arXiv:2506. 06006v3 Announce Type: replace-cross Abstract: Can unified vision-language models (VLMs) perform forward dynamics prediction (FDP), i.

By Yifu Qiu, Yftah Ziser, Anna Korhonen, Shay B. Cohen, Edoardo M. Ponti

llms fine-tuning multimodal benchmarks

arXiv AI

Jul 7

Domain Knowledge-Informed Self-Supervised Representations for Workout Form Assessment

arXiv:2202. 14019v3 Announce Type: replace-cross Abstract: Maintaining proper form while exercising is important for preventing injuries and maximizing muscle mass gains.

By Paritosh Parmar, Amol Gharat, Helge Rhodin

computer-vision fine-tuning

arXiv AI

Aug 7

UniVVT: A Unified End-to-End Framework for High-Fidelity Video Virtual Try-on

arXiv:2608. 05745v1 Announce Type: cross Abstract: Video Virtual Try-On (VVT) synthesizes a video of a person wearing a target garment while preserving identity, motion, and scene dynamics.

By Yushe Cao, Shikun Feng, Fei Shen, Haikuo Peng, Jianqiang Xia, Yiheng Zhu, Dianxi Shi, Chun Yu

llms diffusion computer-vision multimodal benchmarks safety

Related stories

Gravity-Aware Hierarchical Routing for Lightweight SensorLLM on Human Activity Recognition

ActivityNarrated: An Open-Ended Narrative Paradigm for Wearable Human Activity Understanding

TransfHAR: Self-Supervised Wrist Representations for On-Demand Activity Recognition

Can VLMs Predict Future States? Bootstrapping World Models from Inverse Dynamics

Domain Knowledge-Informed Self-Supervised Representations for Workout Form Assessment

UniVVT: A Unified End-to-End Framework for High-Fidelity Video Virtual Try-on