OpenAI shrinks GPT-5.4 for speed and lower costs

OpenAI’s GPT-5.4 mini and nano models cut costs and latency while staying close to flagship performance, giving developers faster AI options for real-time apps without sacrificing core capabilities.
Scientific Research Publishing

A Lightweight Self-Supervised Representation Learning Framework for Depression Risk Profiling from Synthetic Daily Behavioural Trajectories ()

Detecting behavioural signatures of depression from everyday digital traces is a central challenge in computational psychiatry. Real-world datasets from smartphones and wearables often suffer from ...

Apple unveils AI model that creates lifelike 3D objects from just one photo

Apple has introduced a new AI model that can turn a single image into a realistic 3D object with accurate lighting and ...
IEEE

Interactive Vehicle Trajectory Prediction Based on Parameterized Transfer Learning Using Encoder–Decoder Network

Abstract: Vehicle trajectory prediction is important for automated vehicles to understand driving scenarios. This paper proposes an encoder-decoder network-based parameterized transfer learning ...
marktechpost

Google DeepMind Introduces Unified Latents (UL): A Machine Learning Framework that Jointly Regularizes Latents Using a Diffusion Prior and Decoder

Generative AI’s current trajectory relies heavily on Latent Diffusion Models (LDMs) to manage the computational cost of high-resolution synthesis. By compressing data into a lower-dimensional latent ...
IEEE

DTSF-CDNet: A Dual-Branch Encoder-Decoder Network with Differential Transformer Skip Fusion for Image Change Detection

Abstract: Change detection plays a vital role in numerous real-world domains, aiming to accurately identify regions that have changed between two temporally distinct images. Capturing the complex ...