More about HKUST
From Analysis to Engagement: AI-Supported Performance Feedback in Sport
PhD Qualifying Examination
Title: "From Analysis to Engagement: AI-Supported Performance Feedback in Sport"
by
Miss Qiaoyi CHEN
Abstract:
Artificial intelligence has significantly advanced sports performance
analysis, enabling detailed assessments of technique, training load, and
tactical decision-making from video, wearable sensors, and positional data.
However, translating these analytical insights into improved athletic
outcomes depends critically on how athletes and coaches engage with and
reflect on AI-generated feedback: a challenge that sports science and HCI
have pursued with differing methods and evaluation criteria, and limited
cross- disciplinary integration.
To bridge this gap, we present a systematic survey of AI-driven sports
coaching systems through a two-stage analytical framework: Stage 1 examines
how AI systems extract usable performance insights; Stage 2 examines how
athletes and coaches engage with and interpret those insights to support
behaviour change. Following PRISMA guidelines, we searched major HCI and
sports science publication venues, identifying 30 papers organised by
analysis targets and reflection settings.
Our analysis reveals three structural gaps in existing systems: limited
support for longitudinal tracking of athlete development; limited integration
between coach-side analysis and athlete-facing feedback; and limited
consideration of both the athlete's capability profile and the demands of
specific tasks. In addition, we observe a gap in population coverage:
para-athletes, youth athletes, and recreational participants in emerging
sports remain largely underrepresented, and the reference models underlying
current AI systems are primarily built from data that excludes them.
We further discuss how emerging AI capabilities (particularly large language
models, multimodal foundation models, and physiological sensing) create
concrete opportunities to address these gaps, reframing AI sports coaching
from a pipeline of analyses into an adaptive, longitudinal, and socially
situated coaching support system.
Date: Thursday, 2 April 2026
Time: 9:00am - 11:00am
Venue: Room 2132C
Lift 22
Committee Members: Dr. Xiaojuan Ma (Supervisor)
Dr. Arpit Narechania (Chairperson)
Prof. Raymond Wong