Video Description:
Boston Dynamics has unveiled the next evolution of its Atlas AI robot tech with the ability to grip and manipulate objects autonomously in a stride towards manual labor automation. New 3D modeling synthesis artificial intelligence called GeoCode can produce high quality mesh outputs as well as control high level shape editing. Breakthrough text to video (T2V) artificial intelligence called Tune-A-Video uses specialized Sparse-Causal Attention to learn smooth movement, generating videos from text prompts through an efficient one-time adjustment of pre-trained text to image diffusion models.
- Category
- Variety
- Tags
- boston dynamics, ai robot, atlas, a.i., automation, artificial intelligence, robotics, robots, machine learning, deep learning, computer vision, neural network, neural networks, tech news, new tech, new technology, technology news, future tech, the singularity, 3d modeling, blender, reinforcement learning, nlp, ml, natural language processing, openai, open ai, humanoid, transhumanism, cyborg, google, deepmind, microsoft, agi, 3d video, 3d shapes, 3d printing, 3d printer, optimus, tesla
|
-
14:35
-
02:43
-
00:48
-
02:07
-
11:01
-
05:18
-
03:16
-
13:21
-
08:05
-
08:01
-
04:35
-
04:09
-
08:22
-
07:16
-
03:49
-
10:02
-
04:23
Loading...
These are always amazing . . .