I am a 4th-year Ph.D. student in Electrical and Computer Engineering at Carnegie Mellon University (CMU), advised by Prof. Anthony Rowe in WiseLab. My research sits at the intersection of AI, human-computer interaction, extended reality, and spatial intelligence, focused on developing cognitive assistants for the physical world. My work spans multimodal machine learning and LLMs, 3D scene understanding, spatial knowledge representation, and natural language-driven XR program generation, with a focus on physical grounding and personalization. I am passionate about creating human-centered AI systems that provide contextual, spatially-aware assistance for real-world applications.
I have had the opportunity to intern at Dolby Laboratories, where I worked on 4D reconstruction from egocentric devices as part of the Spatial Computing Team in the Advanced Technology Group. I have also interned at Bosch Research, extending XR and LLM-based cognitive assistants for industrial task guidance, and at NVIDIA, building real-time sensor monitoring platforms for autonomous vehicles.
I completed my undergraduate degree from Carnegie Mellon University (CMU) with a Bachelor of Science in Electrical and Computer Engineering and Human Computer Interaction.
Updates
- Oct 2026 — Will be attending and presenting my TVCG paper - 3D Model Representations at ISMAR 2026 in Bari, Italy
- May–Aug 2026 — PhD Research Intern at Dolby Laboratories, Sunnyvale, Spatial Computing Team, Advanced Technology Group
- Apr 2026 — Position paper accepted at CHI 2026 Workshop on Human-AI-UI Interactions Across Modalities in Barcelona, Spain
- Mar 2026 — Presented GenAssist as a talk and demo at IEEE VR 2026 in Daegu, South Korea
Publications
SPIN: Spatial Instruction Authoring from Egocentric Multimodal Data
Sruti Srinidhi, Tao Jin, David Lindlbauer, Anthony Rowe
Under submission at CHI 2027
A system for authoring spatial instructions through in-situ narration. An author wearing smart glasses walks through a space and describes tasks, constraints, and how objects are used, from which SPIN automatically constructs a spatially aware knowledge graph that serves as a queryable representation of the semantics of the space. The system also generates targeted follow-up questions to elicit missing knowledge, yielding a more complete set of instructions.
Learning Visual and Motion Aware 3D Model Representations for Semantic Retrieval
Sruti Srinidhi , Anthony Rowe
TVCG 2026: IEEE Transactions on Visualization and Computer Graphics, presented at ISMAR 2026
A machine learning approach for semantic 3D model retrieval that captures both visual appearance and motion. Each model is represented through multiview renders and short animations, and features from pretrained image–text and video–text encoders are combined by a trained multimodal fusion model into a unified embedding space, enabling retrieval from text, image, and video queries.
GenAssist: Interactive Prompt-Driven XR Program Generation
Sruti Srinidhi , Akul Singh, Edward Lu, Anthony Rowe
IEEE VR 2026: IEEE Conference on Virtual Reality and 3D User Interfaces
A system for generating interactive XR programs from natural language prompts. It uses retrieval-augmented generation over API documentation to synthesize programs, followed by an iterative correction loop in which an LLM inspects visual and textual snapshots of the running program and refines the code.
🏆 XaiR: An XR Platform that Integrates Large Language Models with the Physical World
Sruti Srinidhi , Edward Lu, Anthony Rowe
ISMAR 2024: IEEE International Symposium on Mixed and Augmented Reality
Best Student Paper Award
An XR platform that integrates multimodal LLMs with the physical world. It streams camera, audio, and depth data from a headset to a server, where the data is processed and used to query LLMs, and their responses are projected back into the 3D environment as AR overlays, enabling spatially grounded assistance for real-world tasks.
NFCapsule: An Ingestible Sensor Pill for Eosinophilic Esophagitis Detection Based on near-Field Coupling
Junbo Zhang, Gaurav Balakrishnan, Sruti Srinidhi, Arnav Bhat, Swarun Kumar, and Christopher Bettinger
SenSys '22: Proceedings of the 20th ACM Conference on Embedded Networked Sensor Systems
An ingestible sensor pill for the detection of eosinophilic esophagitis based on near-field coupling.
Clean Vibes: Hand Washing Monitoring Using Structural Vibration Sensing
Jonathon Fagert, Amelie Bonde, Sruti Srinidhi, Sarah Hamilton, Pei Zhang, Hae Young Noh
2022 ACM Transactions on Computing for Healthcare (HEALTH), Volume 3, Issue 3
A system for monitoring hand washing using structural vibration sensing, detecting hand washing activity from the vibrations it induces in the sink.
Workshop Papers
Generative Situated XR Assistance: Generating Context-Aware Guidance for the Physical World (Position Paper)
Sruti Srinidhi, David Lindlbauer, Anthony Rowe
CHI 2026 Workshop on Human-AI-UI Interactions Across Modalities
A position paper proposing Generative Situated XR Assistance, in which systems continuously perceive the user's physical context and generate spatial guidance at runtime rather than relying on pre-authored content.
Demos
An XR Platform that Integrates Large Language Models with the Physical World.
Sruti Srinidhi , Edward Lu, Akul Singh, Saisha Kartik, Audi Lin, Tarana Laroia, Anthony Rowe
SenSys '25: Proceedings of the 23rd ACM Conference on Embedded Networked Sensor Systems
GenAssist: Interactive Prompt-Driven XR Program Generation
Sruti Srinidhi , Akul Singh, Edward Lu, Anthony Rowe
IEEE VR 2026: IEEE Conference on Virtual Reality and 3D User Interfaces
Talks
GenAssist: Interactive Prompt-Driven XR Program Generation
CyLab Student Seminar, 2026
XaiR: An XR Platform that Integrates Large Language Models with the Physical World
Bosch Technical Colloquium (Real Time Distributed Systems), 2025
Building Smart Physical World Assistants: AI-Driven Spatial Guidance and Expressive XR Program Generation
BoCSE – Bosch Internal Conference, Germany 2025
Using Large Language Models to Understand the Physical World
HCI Seminar Talk, Columbia University, 2024
XaiR: An XR Platform that Integrates Large Language Models with the Physical World
AREA Research Webinar, 2024