Mengzhe Geng
PulseAugur coverage of Mengzhe Geng — every cluster mentioning Mengzhe Geng across labs, papers, and developer communities, ranked by signal.
2 day(s) with sentiment data
-
New CIViC protocol rigorously tests audio's influence on visual AI models
Researchers have developed CIViC, a four-gate protocol designed to rigorously evaluate how audio information influences visual prediction models. This method links interpretation to intervention, comparison on held-out …
-
Qwen2-Audio quantization performance evaluated beyond text scores
A new study published on arXiv evaluates the performance of quantized Qwen2-Audio-7B-Instruct models on speech tasks. The research highlights that text-based scores alone are insufficient to determine if quantization pr…
-
Speech deepfake detection improved with auditable evidence
Researchers have developed a novel approach to speech deepfake detection that moves beyond simple scoring to provide auditable evidence for decisions. This method incorporates four aligned cues: a passive detector score…