About
Research & Development
Accounting for Training Randomness in Machine Learning Influence Estimation
TLDR: A new framework called f-INE uses hypothesis testing to reliably estimate the influence of individual data points on machine learning models, even when...
Research & Development
ECO: Guiding Code-LLMs to Write Faster, More Efficient Code
TLDR: ECO (Enhanced Code Optimization) is a new framework that significantly improves the ability of code-LLMs to optimize code for runtime performance. Unlike previous...
Research & Development
BitMar: Bringing Advanced Multimodal AI to Edge Devices
TLDR: BitMar is a new, highly efficient 1.58-bit quantized multimodal AI model designed for edge devices. It combines low-bit text and vision encoders with...
Research & Development
Machine Learning Powers Automated Guitar Tablature Generation from MIDI
TLDR: This paper presents a machine learning method for converting MIDI musical parts into guitar tablature. It uses a deep neural network to generate...
Research & Development
Securing Facial Authenticity with a New Vision Foundation Model
TLDR: FS-VFM is a self-supervised pre-training framework that learns fundamental representations of real faces to detect deepfakes, diffusion forgeries, and face spoofing. It uses...
Research & Development
Real-Time Voice Clarity in Vehicles: Introducing LSZone
TLDR: LSZone is a new lightweight architecture for real-time in-car multi-zone speech separation. It addresses the high computational cost of previous methods by using...