2 papers
cs.SD2025
Cross-Modal Watermarking for Authentic Audio Recovery and Tamper Localization in Synthesized Audiovisual Forgeries
Minyoung Kim, Sehwan Park, Sungmin Cha +1
Recent advances in voice cloning and lip synchronization models have enabled Synthesized Audiovisual Forgeries (SAVFs), where both audio and visuals are manipulated to mimic a targ…
cs.LG2025
Random Conditioning with Distillation for Data-Efficient Diffusion Model Compression
Dohyun Kim, Sehwan Park, Geonhee Han +2
Diffusion models generate high-quality images through progressive denoising but are computationally intensive due to large model sizes and repeated sampling. Knowledge distillation…