Institution profile

Gwangju Institute of Science and Technology

Academic institutionasia · kr
Official website
Research library128linked papers
Opportunities0open roles
Selected work

Representative Papers

PRiSM: Benchmarking Phone Realization in Speech Models

Jan 20, 2026

Current evaluations of phoneme recognition systems are largely confined to surface-level transcription accuracy, with limited insight into their underlying phonemic perception capabilities. This work proposes PRiSM—the first open-source comprehensive benchmark for phoneme recognition—which establishes a standardized evaluation framework integrating both intrinsic (representation probing) and extrinsic (downstream tasks across clinical, educational, and multilingual settings) assessments. Leveraging transcription-based metrics, multilingual datasets, and encoder-CTC architectures, the benchmark enables reproducible evaluation and reveals that multilingual training substantially enhances performance, encoder-CTC models exhibit the most consistent results, and specialized phoneme recognition models still outperform general-purpose large audio language models.

1 citationsRead paper

Automated Taxi Booking Operations for Autonomous Vehicles

Dec 01, 2019International Conference on Signal Processing and Communication Systems

Conventional taxi dispatch paradigms—designed for human drivers—are ill-suited for autonomous vehicles (AVs), posing significant challenges in real-time, driverless order fulfillment. To address this, we propose and deploy the first end-to-end automated ride-hailing system specifically designed for production-level AVs. Our system integrates a mobile client app, a centralized dispatch server, an in-vehicle communication module, and real-time path planning algorithms, enabling seamless request initiation, dynamic vehicle assignment, autonomous acceptance decision-making, and bidirectional human–vehicle state synchronization. Its key innovation lies in unifying full-stack order dispatch, vehicle-level autonomous response, and low-latency human–vehicle coordination on a real-world AV platform—eliminating reliance on remote human operators. Experimental evaluation demonstrates an average response time of <8 seconds, 99.2% order matching accuracy, and robust support for concurrent multi-vehicle/multi-passenger dispatch, thereby validating the system’s technical feasibility and operational practicality.

1 citationsRead paper
Recent publications

Latest Papers