RESEARCH INITIATIVES/Wearable Computing & Memory Augmentation
EgoScore: Importance-Weighted Retrieval for Egocentric Lifelogging via Multi-Signal Perceptual Scoring
Wearable Computing & Memory Augmentation● ACTIVE RESEARCH

EgoScore: Importance-Weighted Retrieval for Egocentric Lifelogging via Multi-Signal Perceptual Scoring

Project Summary

When users wear always-on egocentric cameras (such as Meta Ray-Ban glasses), they accumulate hours of video daily. EgoScore operationalizes moment importance across social significance, emotional salience, novelty, and memorability by fusing six perceptual signals with Ebbinghaus-inspired temporal decay.

RESEARCH OVERVIEW & DETAILS

Project Scope & Details

EgoScore introduces importance-weighted retrieval for egocentric lifelogging. Drawing on Dual-Process Memory Theory (Tulving & Baddeley) and Event Segmentation Theory (Zacks et al.), the system indexes 2-minute egocentric video segments into a steerable topological memory graph.\n\nThe pipeline extracts six orthogonal perceptual signals—SigLIP visual novelty, YOLOv8/v11 social presence, optical flow activity intensity, Whisper audio affect, MediaPipe interaction density, and Mahalanobis contextual surprise—fused with an Ebbinghaus forgetting curve. In a within-subjects user study (N=16) searching personal lifelogs, importance-reranked retrieval achieved significantly higher satisfaction than semantic-only retrieval (p < .01), with 75% of participants preferring the importance-aware system for daily lifelogging and memory augmentation.

ARCHITECTURAL & SCIENTIFIC HIGHLIGHTS
01.

Poincaré Hyperbolic Graph Retrieval

Continuous Riemannian manifold embeddings supporting hierarchical episodic memory recall without geometric distortion.

02.

Multi-Signal Perceptual Scoring

Fuses 6 orthogonal visual-acoustic signals (head fixation, audio energy, face engagement, zoom kinematics, novelty, emotion) into composite frame importance.

03.

Ebbinghaus Temporal Retention

Cognitive exponential decay modeling that organically deprioritizes routine experiences while preserving key autobiographical anchors.

04.

Meta Ray-Ban Wearables Integration

Native Android integration using Meta Wearables DAT for hands-free 1080p egocentric capture and spatial microphone streaming.

INTERACTIVE 3D SPATIAL MODEL & KINEMATIC TELEMETRY

Real-Time 3D Mesh & Spatial Coordinate Simulator

Explore an interactive WebGL OpenXR spatial asset preview. Rotate, orbit, and toggle mesh visualization modes to inspect simulated sub-millimeter 6DOF coordinate tracking.

PROJECT GALLERY [4 PHOTOS]
Click to expand
EgoGraph System Architecture
FIG. 01

EgoGraph System Architecture

Perceptual Signal Extraction Flow
FIG. 02

Perceptual Signal Extraction Flow

Poincaré Hyperbolic vs Euclidean Embedding
FIG. 03

Poincaré Hyperbolic vs Euclidean Embedding

Assistive Memory Prosthetic Architecture
FIG. 04

Assistive Memory Prosthetic Architecture

ACADEMIC PUBLICATIONS [1]
2027Under Review, ACM Conference on Human Factors in Computing Systems (CHI 2027)

EgoScore: Importance-Weighted Retrieval for Egocentric Lifelogging via Multi-Signal Perceptual Scoring

Authors: Chaitanya Anand, Himangshu Sarma

RESEARCH TEAM
  • Chaitanya Anand
  • Dr. Himangshu Sarma
TECHNOLOGIES
Egocentric Computer VisionWearable ComputingPoincaré Hyperbolic EmbeddingsSigLIP ViT-So400MMulti-Signal FusionEbbinghaus Temporal DecayMeta Ray-Ban Smart GlassesFAISS HNSW32Human-Centered AI
PREVIOUS PROJECTDPLA-Retriever: Memory-Efficient Web-Scale Retrieval and Single-GPU LLM Co-location
NEXT PROJECTAlgorithmic Collusion in Two-Sided Labor Markets: AI Resume Screeners, Monoculture, and Wage Compression
HCI Lab IIITS | Human-Computer Interaction Laboratory