AI/ML News

Stay updated with the latest news and articles on artificial intelligence and machine learning

Exploring Qwen3.5 family: from small to massive

Alibaba’s Qwen3.5 combines multimodal intelligence and advanced reasoning with ultra-efficient compute through MoE sparsity and native vision-language fusion. Spanning compact on-device models to massive flagship versions, this open-weight family brings high-performance AI to everything from smartphones to cloud-scale servers.

Cracking the cellular code with APOLLO

APOLLO, a new AI framework, separates shared biological signals across measurement types from those unique to each technique. This unlocks clearer insights into cell states, predicts unmeasured features, spots disease biomarkers more precisely, and could speed up discoveries in cancer, Alzheimer’s, and beyond.

The Self-Taught AI Redefines Computer Vision

Meta AI’s DINOv3 is a self-supervised vision model trained on 1.7 billion images, setting new standards in image classification, object detection, and beyond. With innovations like Gram anchoring and real-world impact from monitoring deforestation to powering NASA’s Mars exploration, it marks a paradigm shift in computer vision.

The rise of the collective machine mind

The new system enables groups of robots to act as a unified team. The MultiRobot FrameWork lets robots share real-time information about their environment, positions, and tasks, mirroring the collective behavior seen in insect colonies, but powered by advanced sensors and computation.

Bridging the data gap in medical imaging with AI

The new GenSeg framework significantly reduces the need for expert-labeled data and achieves high-accuracy medical image segmentation with as few as 40-50 samples. By creating realistic synthetic scans paired with exact labels, it empowers the development of advanced diagnostic tools even in data-limited settings.

AI eye matches human color perception

A self-powered artificial synapse can mimic human color vision with 10-nanometer resolution using dye-sensitized solar cells. This technology enables energy-efficient AI systems capable of advanced color recognition and logic processing.

AI learns to sync sight and sound

MIT researchers have developed CAV-MAE Sync, an AI model that learns to precisely link sounds with matching visuals in video without any labels. This technology can bring us closer to smarter AI that can see, hear, and understand the world just like humans.

AI tool enhances transparency in X-ray analysis

ItpCtrl-AI improves X-ray diagnostics by mimicking radiologists' gaze patterns, providing interpretable heatmaps that enhance transparency and trust in AI-driven medical imaging. By filtering out irrelevant data and focusing on key diagnostic areas, the system ensures more accurate and explainable results.

Autonomous landing innovation – a new era for drones

The Indian Patent Office has granted a patent for the innovative landing system for mini-UAVs. This technology enables precise landings in challenging terrains and has potential applications in both military and civilian logistics, including high-altitude deliveries and emergency.

Collision avoidance system transforms drone navigation

A low-cost, innovative accident avoidance system for drones uses onboard sensors and cameras to autonomously prevent mid-air collisions. This technology is crucial for UAV operations, ensuring safety and efficiency in increasingly crowded airspaces.

Advanced vision system inspired by praying mantis eyes

A new computer vision system significantly reduces energy consumption while providing real-time, realistic spatial awareness. It enhances AI systems' ability to accurately perceive 3D space – crucial for technologies like self-driving cars and UAVs.

MIT's MAIA: an automated agent for interpreting AI models

MAIA can interpret neural networks by conducting experiments and refining its analysis, enhancing understanding of AI models. This agent can identify neuron activities, remove irrelevant features, and detect biases, making AI systems safer and more transparent.

Creating digital elevation models from open data

Nowadays, users can create DEMs with just one click, thanks to radar satellites providing continuous, high-precision data on the Earth's surface and increasingly fast and accessible open-source software. This allows for effective monitoring of terrain changes and natural phenomena.

Zephyr drone is breaking records in the stratosphere

The solar-powered Zephyr drone has set world records for endurance and altitude, staying aloft for 64 days at heights of up to 75,000 feet. With applications ranging from earth observation to mobile phone base stations, Zephyr provides critical connectivity in remote areas.

A new era of multimodal AI with GPT-4o

During the Spring Update event OpenAI’s presented GPT-4о – the unique omnimodel that integrates text, audio and image processing, allowing it to work faster and more efficiently than ever before.

Stable Diffusion 3 – next-gen AI image generator

Stability AI presented the latest advancement in image generative AI models – Stable Diffusion 3. Its expanded parameter range and diffusion transformer architecture ensure smooth generation of complex, high-quality images and accurate text-to-visual translation.

Tracking every pixel: motion estimation with OmniMotion

The latest motion estimation method can extract long-term motion trajectories for every pixel in a frame, even in the case of fast movements and complex scenes. Learn more about the exciting technology and the future of motion analysis in this article about OmniMotion.

Benefits of the Look to Speak

Look to Speak is designed to help those with motor function impairments and speech difficulties to communicate more easily. The app lets people use their eyes to select pre-written phrases and have them spoken out loud.

How sound can model the world

MIT researchers have developed a machine-learning technique that precisely collects and models the underlying acoustics of a location from just a limited number of sound recordings.