News

IBM’s open-source TerraMind AI uses 9 data modalities to transform Earth observation
To build the TerraMesh dataset that underpins TerraMind, IBM’s researchers compiled data on everything from biomes to land ...
marktechpost13d
Decoupled Diffusion Transformers: Accelerating High-Fidelity Image Generation via Semantic-Detail Separation and Encoder Sharing
Researchers from Nanjing University and ByteDance Seed Vision introduce the Decoupled Diffusion Transformer (DDT), which separates the model ... a condition encoder and a velocity decoder to handle ...
AI News This Week from Google, OpenAI, Meta and Anthropic
Uncover the week’s top AI developments, from Google’s AGI push to Anthropic’s Claude updates, and their implications for the ...
the-decoder16d
Researchers introduce COLORBENCH to test color understanding in vision-language models
COLORBENCH assesses models across three main dimensions: color perception, color reasoning, and robustness to color alterations. The benchmark includes 11 tasks with a total of 1,448 instances and ...
marktechpost17d
Meta AI Introduces Perception Encoder: A Large-Scale Vision Encoder that Excels Across Several Vision Tasks for Images and Video
As AI systems grow increasingly multimodal, the role of visual perception models becomes more complex. Vision encoders are expected not ... 3.3. These synthetic annotations allow the same image ...
IEEE18d
Decoder-Only Image Registration
Abstract: In unsupervised medical image registration, encoder-decoder architectures are widely used to predict dense, full-resolution displacement fields from paired images. Despite their popularity, ...
IEEE1d
Driver Drowsiness Detection Using Swin Transformer and Diffusion Models for Robust Image Denoising
This study presents a robust and scalable driver drowsiness detection framework that integrates a Swin Transformer-based deep learning model with a diffusion model for image denoising ... model built ...