Research Scientist Intern, Wearable Sensing & Human Motion Analysis

Meta • Redmond, WA

9d • $7,650 - $12,667

About The Position

The Meta Reality Labs Research Team brings together researchers, developers, and engineers to create future AR and VR technologies. The Surreal Vision group is seeking a Research Scientist Intern to develop egocentric machine-perception capabilities and digital twins of 3D scenes for future VR and contextual-AI-enabled AR devices.

This internship focuses on developing a learned model for swing-sports motion analysis as an alternative to heuristic, threshold-based methods. Goals include improving analysis accuracy, generalization across players and equipment, and producing interpretable feedback. The intern will contribute to egocentric data collection using the team's simulator research setup as well as model development. Internships last 12 to 24 weeks, with various start dates throughout the year; some projects may require a minimum of 24 consecutive weeks.

Requirements

• Currently has, or is in the process of obtaining, a PhD in Computer Science, Electrical Engineering, Robotics, Biomechanics, or a related field.
• Strong Python skills and hands-on experience training models in PyTorch.
• Working knowledge of sequence and time-series modeling, such as temporal CNNs, transformers, state-space models, or HMM/CTC-style segmentation.
• Experience working with real sensor data rather than only curated benchmarks.
• Comfort with conducting hands-on hardware experiments with human participants in a laboratory setting.
• Must obtain work authorization in the country of employment by the time of hire and maintain ongoing work authorization during employment.

Nice To Haves

• Experience designing and executing data collections with wearable or motion-capture hardware.
• Research experience with IMU-based human activity recognition, pose or motion estimation, or wearable biosignals.
• Experience with surface EMG, including filtering, rectification, envelope extraction, electrode shift, and cross-session variability.
• Familiarity with egocentric perception, including Project Aria, Ego4D/Ego-Exo4D, or Aria Digital Twin, or with multimodal sensor fusion.
• Background in sports biomechanics or motion analysis.
• Experience with self-supervised or few-shot learning for personalization.
• First-author publications at CVPR, ICCV, ECCV, NeurIPS, or ICLR, or in biomechanics and wearables venues.
• Plays swing sports, which is helpful for labeling and sanity-checking feedback.
• Intent to return to a degree program after completing the internship.

Responsibilities

• Operate egocentric data collections on an in-lab simulator, instrumenting participants with Aria Gen 2 and Neural Band wearables, running capture sessions, and using simulator ground truth for model supervision and validation.
• Design and train models on multimodal egocentric time-series data, fusing Aria Gen 2 SLAM/IMU, eye gaze, and RGB with Neural Band surface EMG and wrist IMU.
• Develop methods for temporal swing-phase segmentation, kinematic and intent estimation from head-mounted and wrist-worn viewpoints, and swing-quality scoring.
• Build an evaluation harness to benchmark the model against the existing threshold baseline on held-out players, including accuracy, phase-boundary error, and cross-subject generalization.
• Define capture and labeling protocols and address small, noisy, personalized datasets using approaches such as self-supervised pretraining, augmentation, and subject-adaptive fine-tuning.
• Integrate the winning model into the prototype and demonstrate it end to end.
• Write up results; strong work may be considered for publication or an internal technical report.

What This Job Offers

• Career Level: Intern
• Education Level: Ph.D. or professional degree