Update README.md
Browse files
README.md
CHANGED
|
@@ -3,11 +3,8 @@ Welcome to the official Hugging Face home of Circulus. We are a Physical AI comp
|
|
| 3 |
|
| 4 |
🌐 Our Vision: Physical AI for the Real World
|
| 5 |
Circulus moves beyond "Screen-only AI." We focus on Physical AI, where intelligence is embedded into robotic bodies that can perceive, reason, and act in real-time. To achieve this safely and efficiently, we prioritize On-Device AI.
|
| 6 |
-
|
| 7 |
Zero Latency: Real-time processing for natural human-robot interaction.
|
| 8 |
-
|
| 9 |
Privacy-First: Personal data is processed locally, ensuring maximum security.
|
| 10 |
-
|
| 11 |
True Autonomy: Devices function reliably even in offline environments.
|
| 12 |
|
| 13 |
🤝 Strategic Partnership: Intel x Circulus
|
|
@@ -15,32 +12,5 @@ As a key Intel Physical AI Partner, Circulus maximizes the potential of edge har
|
|
| 15 |
|
| 16 |
⚡ Technical Synergy
|
| 17 |
Intel® Core™ Ultra & NPU Optimization: We leverage the latest NPU (Neural Processing Unit) and GPU capabilities to run complex VLMs and LLMs on low-power devices.
|
| 18 |
-
|
| 19 |
OpenVINO™ Integration: Our models are fine-tuned and quantized using the Intel® OpenVINO™ toolkit to ensure peak inference performance.
|
| 20 |
-
|
| 21 |
PiCore (On-device AI Board): Our specialized hardware acts as the "brain" of Physical AI, powered by Intel technology and optimized for multimodal AI workloads.
|
| 22 |
-
|
| 23 |
-
🛠 What We Share on Hugging Face
|
| 24 |
-
We contribute to the community by providing resources that bridge the gap between AI research and robotics deployment:
|
| 25 |
-
|
| 26 |
-
Optimized Edge Models: Quantized versions (INT4, INT8) of SLMs (Small Language Models) and VLMs tuned specifically for Intel AI PCs and edge platforms.
|
| 27 |
-
|
| 28 |
-
Robotics Foundation Models: Lightweight models designed for motion control, vision-language reasoning, and human-robot interaction.
|
| 29 |
-
|
| 30 |
-
Physical AI Datasets: Curated data focusing on real-world robotics environments and multimodal sensor fusion.
|
| 31 |
-
|
| 32 |
-
🚀 Our Ecosystem
|
| 33 |
-
ARCos (AI Robot OS): A specialized operating system that harmonizes CPU, GPU, and NPU workloads for robotics.
|
| 34 |
-
|
| 35 |
-
PIBO & piOn: Our flagship humanoid and companion robots that serve as the ultimate testbeds for our On-device Physical AI solutions.
|
| 36 |
-
|
| 37 |
-
🔗 Connect with Us
|
| 38 |
-
Stay updated with our latest releases in Edge AI and Robotics.
|
| 39 |
-
|
| 40 |
-
Website: circul.us
|
| 41 |
-
|
| 42 |
-
GitHub: Circulus GitHub
|
| 43 |
-
|
| 44 |
-
Contact: [contact@circul.us]
|
| 45 |
-
|
| 46 |
-
"Bridging the gap between code and kinesthetics."
|
|
|
|
| 3 |
|
| 4 |
🌐 Our Vision: Physical AI for the Real World
|
| 5 |
Circulus moves beyond "Screen-only AI." We focus on Physical AI, where intelligence is embedded into robotic bodies that can perceive, reason, and act in real-time. To achieve this safely and efficiently, we prioritize On-Device AI.
|
|
|
|
| 6 |
Zero Latency: Real-time processing for natural human-robot interaction.
|
|
|
|
| 7 |
Privacy-First: Personal data is processed locally, ensuring maximum security.
|
|
|
|
| 8 |
True Autonomy: Devices function reliably even in offline environments.
|
| 9 |
|
| 10 |
🤝 Strategic Partnership: Intel x Circulus
|
|
|
|
| 12 |
|
| 13 |
⚡ Technical Synergy
|
| 14 |
Intel® Core™ Ultra & NPU Optimization: We leverage the latest NPU (Neural Processing Unit) and GPU capabilities to run complex VLMs and LLMs on low-power devices.
|
|
|
|
| 15 |
OpenVINO™ Integration: Our models are fine-tuned and quantized using the Intel® OpenVINO™ toolkit to ensure peak inference performance.
|
|
|
|
| 16 |
PiCore (On-device AI Board): Our specialized hardware acts as the "brain" of Physical AI, powered by Intel technology and optimized for multimodal AI workloads.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|