Dareen Alharthi | Portfolio

I am a PhD student at Carnegie Mellon University's Language Technologies Institute, advised by Professor Bhiksha Raj in the Machine Learning for Signal Processing (MLSP) Lab, where I also earned my Master's in Language Technologies. Before CMU, I received my Bachelor's degree in Computer Science from IMSIU University, where I worked with Professor Areeb Alowisheq. My research focuses on representation learning and generative models for speech and audio. Recently, I've been working on editing tasks, in particular how to scale training on noisy and weakly labeled data and make editing robust to that noise.

News

[Jun 2026] "RIVET: Robust Idempotent Voice Attribute Editing" accepted at INTERSPEECH 2026 (Oral presentation).
D. Alharthi, B. Koduru, R. Singh, B. Raj
[Sep 2026] "Encoder-Decoder Manifold Alignment for Idempotent Generation" accepted at the DynaFront Workshop, NeurIPS 2026.
D. Alharthi, I. Aldarmaki, A. Waheed, S. Semenov, B. Raj
[Jul 2026] "Heard but Not Heeded: Paralinguistic Information Encoding and Loss in Audio-Language Models" accepted at COLM 2026.
B. Koduru, D. Alharthi, R. Singh, B. Raj
[Jun 2026] "Subliminal Prosody Learning: Auxiliary Emotion Supervision Redistributes Affective Representations Across ALM Layers" accepted at the Mechanistic Interpretability Workshop, ICML 2026.
B. Koduru, D. Alharthi, R. Singh, B. Raj
[Dec 2024] "Tessellated Linear Model for Age Prediction from Voice" accepted at IEEE ICASSP 2025 (lecture/oral session, Information Forensics and Security).
D. Alharthi, M. Zamani, B. Raj, R. Singh
[Jul 2024] "Evaluating Speech Synthesis by Training Recognizers on Synthetic Speech" accepted at SynData4GenAI Workshop, INTERSPEECH 2024.
D. Alharthi, R. Sharma, H. Dhamyal, S. Maiti, B. Raj, R. Singh
[Jun 2024] "PAM: Prompting Audio-Language Models for Audio Quality Assessment" accepted at INTERSPEECH 2024.
S. Deshmukh, D. Alharthi, B. Elizalde, H. Gamper, M. Al Ismail, R. Singh, B. Raj, H. Wang
[Mar 2025] Gave a guest lecture on "Transformers" for the Introduction to Deep Learning course at CMU.
Watch the lecture