Virual Oral Presentation Slide by Slide

IEEE/ICACT20260043 Slide.27        [Big Slide]	Oral Presentation
	Now, lets talk about Real-Time Performance. We tested this on an AMD Ryzen 7 laptop with no dedicated GPU. We achieved an average inference speed of 30 FPS, which translates to a latency of about 33ms per frame. This is crucial because real-time feedback is necessary for conversation; anything slower than 15 FPS feels "laggy" and breaks the flow. We have successfully achieved the "Real-Time" benchmark on consumer hardware.

[Go to Next Slide]

Select Voice: Virtual Oral Presentation!!