# SlotFM: A Motion Foundation Model with Slot Attention for Diverse Downstream Tasks

arXiv: https://arxiv.org/abs/2509.21673 · HTML: https://mlsanity.com/paper/2509.21673/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 28 of its 52 references that are in the map.
- First posted on arXiv 25 Sep 2025.

## Builds on (most cited first, 28 of 28)

- [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023)
- [Object-Centric Learning with Slot Attention](https://mlsanity.com/paper/2006.15055/) (2020)
- [High Fidelity Neural Audio Compression](https://mlsanity.com/paper/2210.13438/) (2022)
- [Deep multi-scale video prediction beyond mean square error](https://mlsanity.com/paper/1511.05440/) (2015)
- [Soft-DTW: a Differentiable Loss Function for Time-Series](https://mlsanity.com/paper/1703.01541/) (2017)
- [Multi-task Self-Supervised Learning for Human Activity Detection](https://mlsanity.com/paper/1907.11879/) (2019)
- [Large-scale Training of Foundation Models for Wearable Biosignals](https://mlsanity.com/paper/2312.05409/) (2023)
- [Exploring Contrastive Learning in Human Activity Recognition for Healthcare](https://mlsanity.com/paper/2011.11542/) (2020)
- [Self-supervised Learning for Human Activity Recognition Using 700,000 Person-days of Wearable Data](https://mlsanity.com/paper/2206.02909/) (2022)
- [Mobile Sensor Data Anonymization](https://mlsanity.com/paper/1810.11546/) (2018)
- [PaPaGei: Open Foundation Models for Optical Physiological Signals](https://mlsanity.com/paper/2410.20542/) (2024)
- [Assessing the State of Self-Supervised Human Activity Recognition using Wearables](https://mlsanity.com/paper/2202.12938/) (2022)
- [Scaling Wearable Foundation Models](https://mlsanity.com/paper/2410.13638/) (2024)
- [IMUPoser: Full-Body Pose Estimation using IMUs in Phones, Watches, and Earbuds](https://mlsanity.com/paper/2304.12518/) (2023)
- [A Large-Scale Evaluation of Speech Foundation Models](https://mlsanity.com/paper/2404.09385/) (2024)
- [FOCAL: Contrastive Learning for Multimodal Time-Series Sensing Signals in Factorized Orthogonal Latent Space](https://mlsanity.com/paper/2310.20071/) (2023)
- [WEAR: An Outdoor Sports Dataset for Wearable and Egocentric Activity Recognition](https://mlsanity.com/paper/2304.05088/) (2023)
- [Toward Foundation Model for Multivariate Wearable Sensing of Physiological Signals](https://mlsanity.com/paper/2412.09758/) (2024)
- [REBAR: Retrieval-Based Reconstruction for Time-series Contrastive Learning](https://mlsanity.com/paper/2311.00519/) (2023)
- [RelCon: Relative Contrastive Learning for a Motion Foundation Model for Wearable Data](https://mlsanity.com/paper/2411.18822/) (2024)
- [CroSSL: Cross-modal Self-Supervised Learning for Time-series through Latent Masking](https://mlsanity.com/paper/2307.16847/) (2023)
- [CAPTURE-24: A large dataset of wrist-worn activity tracker data collected in the wild for human activity recognition](https://mlsanity.com/paper/2402.19229/) (2024)
- [Beyond Sensor Data: Foundation Models of Behavioral Data from Wearables Improve Health Predictions](https://mlsanity.com/paper/2507.00191/) (2025)
- [Wearable Accelerometer Foundation Models for Health via Knowledge Distillation](https://mlsanity.com/paper/2412.11276/) (2024)
- [Hang-Time HAR: A Benchmark Dataset for Basketball Activity Recognition using Wrist-Worn Inertial Sensors](https://mlsanity.com/paper/2305.13124/) (2023)
- [PRIMUS: Pretraining IMU Encoders with Multimodal Self-Supervision](https://mlsanity.com/paper/2411.15127/) (2024)
- [A multi-sensor human gait dataset captured through an optical system and inertial measurement units](https://mlsanity.com/paper/2111.15044/) (2021)
- [Silent Impact: Tracking Tennis Shots from the Passive Arm](https://mlsanity.com/paper/2507.23215/) (2025)

## Built on by (most cited first, 2 of 2)

- [MorphCL: Morphological Contrastive Learning for Inertial-based Human Activity Recognition](https://mlsanity.com/paper/2610.10245/) (2026)
- [IMU-Bench: Benchmarking Wrist IMU Representations Across Kinematic, Semantic and Clinical Tasks](https://mlsanity.com/paper/2610.19481/) (2026)

## Abstract (arXiv)

Wearable accelerometers are used for a wide range of applications, such as gesture recognition, gait analysis, and sports monitoring. Yet most existing foundation models focus primarily on classifying common daily activities such as locomotion and exercise, limiting their applicability to the broader range of tasks that rely on other signal characteristics. We present SlotFM, an accelerometer foundation model that generalizes across diverse downstream tasks. SlotFM uses Time-Frequency Slot Attention, an extension of Slot Attention that processes both time and frequency representations of the raw signals. It generates multiple small embeddings (slots), each capturing different signal components, enabling task-specific heads to focus on the most relevant parts of the data. We also introduce two loss regularizers that capture local structure and frequency patterns, which improve reconstruction of fine-grained details and helps the embeddings preserve task-relevant information. We evaluate SlotFM on 16 classification and regression downstream tasks that extend beyond standard human activity recognition. It outperforms existing self-supervised approaches on 13 of these tasks and achieves comparable results to the best performing approaches on the remaining tasks. On average, our method yields a 4.5% performance gain, demonstrating strong generalization for sensing foundation models.