VoiceID

88 acoustic features is a serious signal — most detection tools use 10-20 and call it done. Curious which feature clusters are most discriminative in practice: prosody patterns, spectral artifacts, or something else entirely? Also: how does it hold up against voice cloning models that specifically train to reduce synthetic artifacts? That's the arms race that matters.

can't I use truecaller?

VoiceID is a very practical and timely product, especially now that AI-generated audio is becoming more common. The platform is straightforward to use, just upload an audio file or record live audio, and you get results within seconds. What I found most valuable is that it doesn’t just give a simple yes/no answer, but also provides confidence scores, which makes the analysis feel more transparent and useful. Overall, it’s a smart tool for anyone working with media, recruiting, security, customer support, or content verification who needs a quick way to distinguish between human and AI-generated voices. Clean execution and a very relevant use case.

With the rise of AI voice cloning scams, a tool like VoiceID is becoming essential for personal and corporate security. The live voice recording feature for instant detection is a great addition for verifying suspicious calls.

Ai generated audio and detect ai generated is nice feature

88 acoustic features!!!

The 88 acoustic features approach is interesting — I'd love to know how it handles voice cloning tools that specifically target detection evasion. Also, no login required is a great UX choice for a tool people might use in sensitive situations.