Dilsheen Kaur

About

Dilsheen Kaur is a researcher at the forefront of affective computing and privacy-preserving machine learning, with a primary focus on speech emotion recognition (SER). Her work addresses the critical challenge of extracting salient emotional features from human speech while safeguarding user privacy. In her highly cited 2024 paper, "Learning Salient Features for Speech Emotion Recognition using Attention-based Residual Bidirectional LSTM with Federated Learning," Kaur introduces a novel architecture that combines attention mechanisms with residual bidirectional long short-term memory networks. This approach enables the model to focus on the most emotionally expressive segments of speech—such as vocal tone, pitch variations, and speech rhythm—while federated learning ensures that raw audio data never leaves the user's device, mitigating privacy risks. By integrating these techniques, Kaur’s work achieves robust emotion classification without compromising data security, a significant contribution given the sensitivity of biometric data. Her research has garnered attention for its practical implications in human-computer interaction, mental health monitoring, and secure voice-based systems. Kaur’s innovative fusion of deep learning and privacy-preserving methods positions her as a rising voice in ethical AI, with her paper already cited by peers exploring similar intersections of emotion AI and decentralized learning.

Research Focus

Key Achievements

Top Papers

1. 1
   Learning Salient Features for Speech Emotion Recognition using Attention-based Residual Bidirectional LSTM with Federated Learning

   1 citations · 2024

Key Collaborators

Contact & Links