About
Research & Development
Understanding Why Monocular Depth Estimation Models See What They See
TLDR: This research explores how to understand the decision-making process of Monocular Depth Estimation (MDE) models, which predict depth from single images. The study...
Research & Development
Enhancing Reinforcement Learning with Adaptive Demonstration Guidance
TLDR: A new framework, Smooth Policy Regularisation from Demonstrations (SPReD), improves reinforcement learning by using ensemble methods to quantify uncertainty in Q-value estimates. Unlike...
Research & Development
EmoHeal: Crafting Personalized Music Journeys for Emotional Well-being
TLDR: EmoHeal is an AI system that provides personalized therapeutic music by detecting 27 fine-grained emotions from user text, mapping them to musical parameters...
Research & Development
Achieving Clearer Object Edges in AI Depth Perception
TLDR: A new self-supervised method for monocular depth estimation models per-pixel depth as a mixture distribution to achieve significantly sharper object boundaries. By capturing...
Research & Development
Personalized Voice Cloning Through Federated Identity-Style Adaptation
TLDR: FED-PISA is a novel federated learning framework for voice cloning that addresses high communication costs and insufficient personalization in existing methods. It introduces...
Research & Development
Advancing Spoken Language Assessment with a Unified Multimodal AI Model
TLDR: Researchers have developed a new AI model for Spoken Language Assessment (SLA) that evaluates a learner's oral proficiency at the session level, rather...