aiqbench analysis
This paper presents EmotionAI, a fully local computational intelligence pipeline that integrates speech emotion recognition with generative reasoning to perform privacy-preserving conversational analysis without sending audio data off-device. While the system demonstrates functional viability on CPU hardware for specific tasks like interview review, its current accuracy in zero-shot cross-corpus transfer remains significantly lower than traditional domain-specific baselines. The core contribution lies not in achieving state-of-the-art recognition metrics but in establishing an auditable framework that balances imperfect affective evidence with strict data privacy constraints.
EmotionAI: A Privacy-Preserving Computational Intelligence Pipeline for Speech-Emotion-Grounded Conversational Analysis
arXiv:2606.24941v2 Announce Type: replace-cross Abstract: Reviewing recorded interviews for affective cues such as composure and agitation is slow and subjective, and cloud services that could automate the task require sensitive audio to leave the device. EmotionAI is a fully local Computational Intelligence (CI) pipeline that couples Speech Emotion Recognition (SER) with generative reasoning. Speaker diarisation, Whisper Automatic Speech Recognition (ASR) and a wav2vec2 emotion classifier produce per-segment affective evidence, and an adversarial three-model local Large Language Model (LLM) panel turns that evidence into timestamp-grounded, citation-constrained answers. Zero-shot evaluation on the RAVDESS four-class English subset (n = 672) measures the cost of cross-corpus transfer: the deployed classifier scores 48.8% accuracy, above random (24.9%) and majority (28.6%) baselines but below an in-domain MFCC + logistic-regression comparator (71.0%). The complete pipeline runs in a mean 157 s on CPU (real-time factor approximately 1.33) with zero external calls. The contribution is not state-of-the-art SER but an auditable, privacy-preserving integration of imperfect affective evidence into grounded conversational analysis.
Abstract:Reviewing recorded interviews for affective cues such as composure and agitation is slow and subjective, and cloud services that could automate the task require sensitive audio to leave the device. EmotionAI is a fully local Computational Intelligence (CI) pipeline that couples Speech Emotion Recognition (SER) with generative reasoning. Speaker diarisation, Whisper Automatic Speech Recognition (ASR) and a wav2vec2 emotion classifier produce per-segment affective evidence, and an adversarial three-model local Large Language Model (LLM) panel turns that evidence into timestamp-grounded, citation-constrained answers. Zero-shot evaluation on the RAVDESS four-class English subset (n = 672) measures the cost of cross-corpus transfer: the deployed classifier scores 48.8% accuracy, above random (24.9%) and majority (28.6%) baselines but below an in-domain MFCC + logistic-regression comparator (71.0%). The complete pipeline runs in a mean 157 s on CPU (real-time factor approximately 1.33) with zero external calls. The contribution is not state-of-the-art SER but an auditable, privacy-preserving integration of imperfect affective evidence into grounded conversational analysis.
12 pages, 4 figures. Submitted to UK Workshop on Computational Intelligence (UKCI 2026)
Sound (cs.SD); Artificial Intelligence (cs.AI)
arXiv:2606.24941 [cs.SD]
(or arXiv:2606.24941v2 [cs.SD] for this version)
https://doi.org/10.48550/arXiv.2606.24941 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 265 words
Related coverage
[AINews] not much happened today
Latent Space · Sep 10, 2026
NVIDIA GTC Taipei at COMPUTEX: Live Updates on What’s Next in AI
NVIDIA · May 21, 2026
dots.tts.edit: Precisely Controlled Speech Editing with a Continuous Autoregressive Model
arXiv · Aug 6, 2026
NVIDIA Nemotron 3 Nano Omni Powers Multimodal Agent Reasoning in a Single Efficient Open Model
NVIDIA · Apr 28, 2026