AI-Powered Mobile Platform for Democratizing Sports Talent Assessment: An Offline-First, Tamper-Evident Edge-AI Approach

India has an enormous pool of sporting talent in its rural and semi-urban regions, yet most of it is never discovered because talent identification still depends on physical trials at fixed venues, costly equipment and the subjective judgement of coaches. Existing digital alternatives require high- end phones or continuous internet access and are easily defeated by tampered or pre-recorded videos. This paper presents an AI- powered mobile platform that turns an ordinary smartphone into an objective, offline-first sports talent scout. The application records a fitness drill on video, runs MediaPipe pose estimation and quantized TensorFlow Lite / CoreML scoring models directly on the device, and computes repetitions, joint angles and performance metrics without any network connection. A trust engine cross-checks the video against the phone’s inertial sensors (IMU), inspects frame integrity and stores a SHA-256 timestamped hash of every result to deter fraud. Results are kept in an encrypted SQLite database and synchronised through Prisma ORM to a PostgreSQL/MongoDB cloud store when connectivity returns, where a Chroma DB vector index and a web dashboard let coaches, scouts and Sports Authority of India (SAI) officials filter, rank and compare verified athletes. A working prototype demonstrates an overall AI score and a six-axis kinematic profile for an analysed video. The design is projected to cut scouting logistics and data- transfer costs substantially while giving every athlete the same, transparent assessment.

Authors

Institutions

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-18
DOI
https://doi.org/10.5281/zenodo.24346513
Primary Topic
Human Pose and Action Recognition
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

AI-Powered Mobile Platform for Democratizing Sports Talent Assessment: An Offline-First, Tamper-Evident Edge-AI Approach

Abhishek Verma, Aditya Raj, Rakesh Raushan, Adin Fatma
1 links
Zenodo (CERN European Organization for Nuclear Research)
Human Pose and Action Recognition
article

AI-Powered Mobile Platform for Democratizing Sports Talent Assessment: An Offline-First, Tamper-Evident Edge-AI Approach

Abhishek Verma, Aditya Raj, Rakesh Raushan, Adin Fatma
article en

Abstract

India has an enormous pool of sporting talent in its rural and semi-urban regions, yet most of it is never discovered because talent identification still depends on physical trials at fixed venues, costly equipment and the subjective judgement of coaches. Existing digital alternatives require high- end phones or continuous internet access and are easily defeated by tampered or pre-recorded videos. This paper presents an AI- powered mobile platform that turns an ordinary smartphone into an objective, offline-first sports talent scout. The application records a fitness drill on video, runs MediaPipe pose estimation and quantized TensorFlow Lite / CoreML scoring models directly on the device, and computes repetitions, joint angles and performance metrics without any network connection. A trust engine cross-checks the video against the phone’s inertial sensors (IMU), inspects frame integrity and stores a SHA-256 timestamped hash of every result to deter fraud. Results are kept in an encrypted SQLite database and synchronised through Prisma ORM to a PostgreSQL/MongoDB cloud store when connectivity returns, where a Chroma DB vector index and a web dashboard let coaches, scouts and Sports Authority of India (SAI) officials filter, rank and compare verified athletes. A working prototype demonstrates an overall AI score and a six-axis kinematic profile for an analysed video. The design is projected to cut scouting logistics and data- transfer costs substantially while giving every athlete the same, transparent assessment.

Zenodo (CERN European Organization for Nuclear Research)
Noida International University (IN), Greater Noida Institute of Technology (MBA Institute) (IN)
Human Pose and Action Recognition
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon