Nvidia has lost it's moat 🤯
Microsoft has open-sourced a 1-bit inference framework that runs massive 100B parameter models directly on your CPU without GPUs.
82% less energy. 6x faster inference. 100% open-source.
A motion capture rig at a Hollywood studio costs about as much as a house.
Someone built the same thing and gave it away for free on Github.
Its called FreeMoCap.
Grab two or more regular webcams, GoPros, or even your phone. Wave a printed checkerboard in front of them once to calibrate the space. Record yourself moving.
The software turns that footage into full 3D motion data you can drop straight into Blender or export as a spreadsheet.
You don't need a suit or markers or a studio.
Just cameras you probably already own and computer vision doing what used to need a warehouse full of hardware.
https://t.co/z33jJHFP0N
This robot has no front. No back. No wrong direction. 🤯
Meet Argus a 20-legged spherical robot from Duke University designed to move and see in virtually every direction.
Each leg carries a depth camera, giving the robot a near-360° view while its unusual geometry lets it navigate forests, sand and rough terrain.
Robots don’t always need to look human sometimes, they need to break the rules.
🎥 Media: @DukeU / Duke General Robotics Lab
⚠️ This content is shared for informational purposes only. CTO Robotics Media is a media platform and does not own or develop the technology shown. Credit belongs to the original creators.
#Robotics #Robot #AI #Engineering #Innovation #DukeUniversity
Un programador solo quería un reproductor de MP3 web sin anuncios.
Terminó creando algo que parece sacado de una película de espías:
Transferencia de archivos 100% air gapped.
Sin WiFi.
Sin Bluetooth.
Sin estar en la misma red.
Solo luz y códigos QR.
Un teléfono muestra QR que parpadean a toda velocidad.
El otro los captura con la cámara y reconstruye el archivo completo.
Usó Claude y en una sola noche tuvo el prototipo funcionando al 100%.
La IA está convirtiendo a un solo developer en un laboratorio entero.
El futuro no espera. Ya está aquí.
fall-detection-vison
Elderly Fall Detection System
A real-time video-based fall detection system that uses skeleton pose estimation and temporal deep learning to detect when a person falls and stays on the ground. Designed for elderly care facilities, hospitals, and smart home monitoring.