FujiView: Multimodal Late-Fusion for Predicting Scenic Visibility
IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) 2026
PhD Student, Computer Science · AICIP Lab, UT Knoxville
Bryce works on multimodal learning and human-centered AI — teaching machines to fuse camera imagery, weather, and sensor data into forecasts people can act on. He leads FujiView, the lab's scenic-visibility forecasting project, published at WACV 2026.
Bryce Bible is a PhD student in computer science at the University of Tennessee, Knoxville, advised by Dr. Hairong Qi in the AICIP Lab. His research sits where computer vision meets the messiness of the real world: fusing webcam imagery, weather data, and physiological signals into systems that forecast, adapt, and assist.
He leads FujiView, a multimodal forecasting system that predicts whether Mount Fuji will be visible from a given viewpoint — built on more than 100,000 human-labeled webcam images from 40+ cameras paired with concurrent and forecast weather data. The work appeared in the main proceedings of WACV 2026 and reaches roughly 89% same-day and 84% next-day accuracy across five visibility classes. A public web app and dataset release are in progress at fujiview.app.
His second thread is AI for dementia care. Within the MUSICARE-VR collaboration he designed the “Conductor AI” architecture that adapts a virtual-reality music intervention in real time — reading biometrics and activity state to adjust tempo, session pacing, and comfort. He is a co-author on related work integrating reinforcement learning and LLM-based agents for adaptive robotic assistance, and has built interaction apps for the SoftBank Pepper robot used in public demonstrations.
Before the PhD, his master’s work produced the EASIER Air Alert Platform — a distributed indoor air-quality monitoring and alert system deployed toward at-risk elder homes as part of a TVA Connected Communities environmental-justice pilot. He was the platform’s system architect and first author of its CHASE ‘23 paper.
Bryce’s research has a deliberate Japanese thread. He speaks advanced Japanese (N1-level coursework), studied at Kansai Gaidai University, and in summer 2024 organized and led a TLSAMP-funded healthcare-AI research trip to Japan — mentoring two undergraduates through visits to RIKEN, Hokkaido University, Kobe University, and Konan University. FujiView grew out of the same orientation: a vision problem that only exists because people care, deeply, whether they will see the mountain tomorrow.
Bryce has been shipping software since long before graduate school — including a stack of Android apps with several hundred thousand combined downloads and virtual-reality worlds built for social VR platforms. He is the founder of Tanuki Inc., through which FujiView is being developed toward tourism-facing deployments. He also speaks Esperanto, which comes up more often than you’d think.
IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) 2026
arXiv 2025
Alzheimer's & Dementia (AAIC abstract) 2024
Abstract presented at the Alzheimer's Association International Conference 2024.
Master's thesis, University of Tennessee, Knoxville 2024
ACM/IEEE Conference on Connected Health: Applications, Systems and Engineering Technologies (CHASE) 2023