computer vision - Tech News from Simmons Systems

Tech News from Simmons Systems

computer vision

A collection of 198 posts

Meta Unveils Chameleon, a Native Multimodal AI Competitor

Meta introduces Chameleon, a new family of AI models built from the ground up to natively process and generate mixed image and text sequences. The early-fusion architecture achieves state-of-the-art results on visual tasks while remaining competitive in text-only benchmarks.

Apple Researchers Unveil State-of-the-Art Multimodal AI Model MM1

Apple researchers publish a new paper demonstrating breakthroughs in training AI on both text and images. The MM1 model achieves top performance as the company invests heavily to catch up in the generative AI race.

Apple Researchers Unveil MM1 Multimodal AI Model in Billion-Dollar Push

Apple researchers publish a breakthrough paper showing how combining text and image data creates highly capable multimodal AI systems. The MM1 model development reflects Apple's massive $1 billion annual investment to compete in the generative AI race.

Apple Researchers Detail MM1, a State-of-the-Art Multimodal AI Model

A new research paper reveals key architectural and data strategies for building highly capable multimodal large language models, culminating in the MM1 family of AI models.

Google Unveils Imagen: A Photorealistic Text-to-Image AI Model

Google Research introduces Imagen, a new text-to-image diffusion model that combines large language models with high-fidelity image generation to achieve unprecedented photorealism. The model outperforms competitors in human evaluations and sets a new state-of-the-art benchmark.

NVIDIA Unveils StyleGAN Architecture to Revolutionize AI Image Synthesis

NVIDIA Research introduces StyleGAN, a new generator architecture that separates high-level attributes from stochastic variations to produce highly realistic images. The system offers unprecedented control over image features by adjusting styles at each convolution layer.

boston dynamics

Boston Dynamics Atlas Robot Shows Off Impressive Parkour Skills

Boston Dynamics releases a new video showcasing its Atlas robot successfully navigating obstacles and jumping like a parkour expert using real-time computer vision.

artificial intelligence

Major AI Facial Recognition Systems Show Significant Gender and Skin Tone Biases

A new MIT and Stanford study reveals that commercial facial-analysis systems from major tech companies exhibit severe error rate disparities, failing frequently for darker-skinned women while remaining highly accurate for light-skinned men.

Figure Unveils Third-Generation Humanoid Robot Built for Mass Production

Figure reveals Figure 03, a completely redesigned humanoid robot engineered specifically for their Helix AI system, home environments, and high-volume manufacturing. The new model features major upgrades to its vision system, tactile hands, and overall safety.

Stability AI Unveils Stable Virtual Camera for Easy 3D Video Generation

Stability AI releases Stable Virtual Camera, a new multi-view diffusion model that turns single 2D images into immersive 3D videos. The tool offers precise 3D camera controls and is currently available for non-commercial research use.

Anthropic's Claude 3 Opus Vision API Tested on Real-World Tasks

The Roboflow team tests Anthropic's new Claude 3 Opus vision API across various computer vision tasks, finding strong OCR capabilities but notable limitations in object detection and visual question answering.

New Research Paper Analyzes OpenAI's Sora Video Generation Model

A comprehensive review explores the underlying technology, applications, and limitations of OpenAI's text-to-video AI model Sora. The study highlights both the potential impact across various industries and the safety challenges that remain.

Google Unveils Gemini, a Powerful Family of Multimodal AI Models

Google introduces Gemini, a new family of highly capable multimodal models designed to seamlessly understand and combine text, images, audio, and video. The research demonstrates significant advancements in reasoning and cross-modal comprehension across various model sizes.

Google Unveils PaLM-E to Bridge Language Models and Robotics

Google introduces PaLM-E, a new embodied multimodal language model that directly ingests robot sensor data to control machines while retaining advanced visual and language capabilities.

New Machine Learning Model Helps Decode James Webb Telescope Images

A machine learning model named Morpheus is helping scientists classify every pixel of high-resolution images captured by the James Webb Space Telescope. The tool identifies stars and galaxies by adapting computer vision algorithms for astronomical data.