Papers by Hyeonsang Kim
WIGVO: Real-Time Bidirectional Speech Translation over Legacy PSTN Calls via Dual-Session Echo Gating (2026.acl-demo)
Copied to clipboard
| Challenge: | Existing systems assume wideband audio and client-side acoustic echo cancellation (AEC) . a system that enables bidirectional speech translation over the PSTN requires app installation or carrier integration. |
| Approach: | They propose a server-side relay system that enables bidirectional LLM-based speech translation over ordinary telephone calls without app installation or carrier integration. |
| Outcome: | The proposed system is deployed over the public switched telephone network (PSTN) it observes 555 ms median latency, 2,684 mS median translation loops, and COMET semantic adequacy of 0.71 (enko) and 0.62 (koen) |