cv

Ph.D. in Experimental Psychology

• Advisor: Bria Long

B.S. in Cognitive Science

• Specialized in Machine Learning & Neural Computation
• Minor: Computer Science

Graduate Student

• What objects children see, and how that differs from curated datasets. Object frequency and representational geometry across 868+ hours of egocentric video (BabyView, 31 families) compared against THINGS. Methods: YOLOE detection with CLIP-based filtering, CLIP and DINOv3 embeddings, representational similarity analysis. Output: first-author manuscript under review; public reproducibility capsule.
• Why some categories are visually harder than others. Global and local dispersion measures over 85 categories separate viewpoint-driven from instance-driven variability; rankings partially replicate across backbones and are only weakly coupled to frequency. Output: CCN 2026 spotlight; reproducible analysis package and layer-wise embedding explorer.
• Whether children and models converge on 3D shape. MOCHI Kids, a developmental (ages 4–6) adaptation of the MOCHI shape-oddity benchmark, built and deployed for direct human-model comparison on multiview object consistency. Methods: TypeScript and jsPsych client, Express and MongoDB backend, end-to-end tests in CI. Status: piloting.
• How language and vision align in everyday input. Alignment between child-directed speech and co-occurring visual scenes in naturalistic video, using multimodal models as the measure. Output: co-first-author preprint.
• What changes when children act on objects. Identifying viewed and manipulated objects by combining monocular depth estimation, IMU-derived head direction, and hand tracking; building a synchronized capture rig (Pupil Labs Neon eye trackers, two depth cameras, room-view camera) for caregiver-child interaction studies. Status: in progress.

Lab Technician

• Multimodal measurement of infant-caregiver interaction. Ran home-based sessions with 12–36 month infants using head-mounted eye trackers, third-person cameras, IMUs, and lapel microphones. Methods: Python preprocessing pipelines across devices; MATLAB codebase to analyze and visualize multi-stream data; YOLOv8 fine-tuning for infant and parent views; MediaPipe and OpenPose for hand and face detection; WhisperX for transcription and diarization; skeletal tracking for dyadic play; 3D environments in Unity, Blender, and Matterport.

Research Assistant

• How people communicate abstract physics knowledge. Honors thesis. Methods: web-based experiments in JavaScript and Node.js; NLP and manual annotation of 480+ participant responses.

Research Assistant

• How verbal reduplication shapes event conceptualization in Mandarin. Methods: web-based psycholinguistic experiments; development of a pressure-sensing pen for physiological response recording.

Variability in young children's everyday visual experiences of object categories

• Spotlight poster: Cognitive Computational Neuroscience.

Characterizing the inputs to infants' object category representations

• Talks: Beyond Curated Datasets (CCN satellite); Vision Science Society; Southern California Meeting for Investigations in Developmental Science; Egocentric Video Workshop, Stanford.

Quantifying infants' everyday experiences with objects in a large corpus of egocentric videos

• Talks: BabyView Workshop (Understanding Infant Experience Through Egocentric Data); BabyView and 1kD joint lightning talks. Posters: Cognitive Computational Neuroscience; Southern California Meeting for Investigations in Developmental Science.

Learning semantic knowledge based on infant real-time attention and parent in-situ speech

• Talks: Cognitive Science Society (46th Annual Meeting); International Congress of Infant Studies.

Using manual actions to create visual saliency

• Talk: Cognitive Science Society (45th Annual Meeting). Posters: COSMOS; Workshop on Natural Environments, Tasks and Intelligence.

Communicating understanding of physical dynamics in natural language

• Talk: UCSD 35th Annual Undergraduate Research Conference. Poster: Cognitive Science Society (44th Annual Meeting).

PSYC 201B Quantitative Methods in Psychology II

PSYC 201A Statistical Computing and Inference from Data I

PSYC 168 Psychological Disorders of Childhood

PSYC 181 Drugs and Behaviors

PSYC 102 Sensory Neuroscience

PSY 371M Introduction to Machine Learning

COGS 189 Brain Computer Interfaces

COGS 101C Language

UC San Diego

• Dora Deng, Mira Mateo, Juan Pablo Bello, Rochelle Boegeman, Emma Nishikawa

UT Austin

• Elton Martinez, Anagha Kenikar, Jacob Rivera, Ruchi Shah

Co-organizer, Beyond Curated Datasets

Colloquium Committee Representative

• National Eye Institute Early Career Scientist Travel Grant, Vision Science Society

• Anderson Travel & Research Award, UC San Diego

• Norman Henry Anderson Fellowship, UC San Diego

• Triton Research and Experiential Learning Scholars Award, UC San Diego
• Distinction in Cognitive Science, UC San Diego

• Provost's Honors, UC San Diego

• HackSC 1st Place in Entrepreneurship, USC