Weighted Finite State Automata and Linear Algebra - JSALT 2023 Team Presentation
Explore weighted finite state automata and linear algebra in speech and language processing through a comprehensive JSALT 2023 team presentation by Lucas Ondel Yang.
Building a Local Voice Assistant Like Alexa Using LLMs and Python
Learn to build a voice-controlled AI assistant using Python, Streamlit, LangChain, and Ollama in under 100 lines of code. Create your own local version of Alexa through hands-on coding.
The Future of Human-Computer Interaction - Opening Keynote UIST 2024
Discover groundbreaking insights into user interface technology and software development through Prof. Raj Reddy's keynote address at the 37th ACM Symposium on User Interface Software and Technology.
DrawTalking - Building Interactive Worlds by Sketching and Speaking
Discover how to create interactive virtual environments through an innovative combination of hand-drawn sketches and voice commands, revolutionizing digital world-building techniques.
Towards Music-Aware Virtual Assistants
Explore the future of AI technology through innovative research on developing virtual assistants that can understand and interact with musical elements and contexts.
Breaking Down Language Barriers with AI Translation - Session 107
Discover how to implement real-time AI translation capabilities across text, speech, and images in your applications, with deployment options for both cloud and edge devices.
Unveiling New Real-Time Speech and Video Translation Innovations - BRK109
Discover Azure AI's latest real-time speech and video translation capabilities, custom avatars, and voices that help enterprises break language barriers and enhance global communication.
Azure AI Content Understanding - A New Multimodal Service - Session 111
Discover Azure AI Content Understanding's capabilities for processing multimodal data, featuring real-world applications from WPP and Numonix for enhanced enterprise AI solutions and customer experiences.
Mixture of Transformers for Multi-modal Foundation Models
Explore Meta's Mixture-of-Transformers (MoT) architecture, a revolutionary approach to handling multi-modal AI tasks combining text, speech, images, and videos with enhanced efficiency and performance.
Multimodal AI: Understanding Large Language Models with Vision and Audio Capabilities
Explore three key approaches to expanding Language Models beyond text, enabling them to process visual and auditory information, with practical demonstrations using LLaMA 3.2.
Harnessing LLMs as Speech Interfaces for Assistive Robotics
Explore how Large Language Models can revolutionize speech interfaces in assistive robotics, enabling more natural human-robot interactions for accessibility applications.