Meta Unveils SAM 2: A Unified Model for Real-Time Object Segmentation in Images and Videos
The model’s ability to handle zero-shot generalization allows it to segment objects in unseen visual domains without custom adaptation. Additionally, the SA-V dataset’s extensive annotations can aid in training more robust computer vision systems.
Shutterstock and Getty Images Unveil Updated AI Generators Powered by NVIDIA Edify
Shutterstock has launched a new enterprise option for their "Generative 3D" service in commercial beta. Their tool allows designers to quickly create 3D assets and 360-degree HDR environments using text or image prompts. The company says it has been working with NVIDIA for a year to train
