About
Research & Development
The Disconnect: Why LLM Confidence Doesn’t Equal Competence
TLDR: A new research paper reveals that large language models (LLMs) possess a 'solvability belief' (an internal assessment of success) that is distinct and...
Research & Development
Real-Time Codec Avatars in VR: ESCA’s Dual Approach to Performance and Quality
TLDR: ESCA is a new framework that combines advanced post-training quantization techniques with a custom hardware accelerator to enable high-fidelity, real-time photorealistic codec avatars...
Research & Development
SwiftEmbed: Achieving Real-Time Text Embeddings with Unprecedented Speed
TLDR: SwiftEmbed introduces a novel static token lookup methodology for text embedding generation, achieving ultra-fast 1.12ms latency and 50,000 requests per second throughput. This...
Research & Development
Advancing PET Imaging: A CT-Free Approach to Attenuation Correction
TLDR: Researchers have developed a deep learning method that uses a Multiview Ensemble Conditional Diffusion Model to generate synthetic CT images (pseudo-CTs) directly from...
Research & Development
DualCap: Generating Richer Image Captions Through Dual Retrieval
TLDR: DualCap is a lightweight image captioning model that uses a novel dual retrieval system. It combines traditional text-based prompts with new visual prompts...
Research & Development
Unpacking AI’s Approach to Detecting Fake Video News
TLDR: This research introduces MVFNDB, a new benchmark for evaluating multimodal large language models (MLLMs) in video fake news detection. Unlike previous benchmarks, MVFNDB...