v3.0.1 Stable Release

We Give Robots
Vision

The open-source vision framework for edge devices. Runs DeepStream pipelines, YOLO detection, and world models at 60 FPS on NVIDIA Jetson, Intel NPU, and Hailo.

Initialize Engine View Repository

openeyes-engine — bash

$ python -m src.main --camera 0 --enable-face --enable-gesture --enable-pose

[INFO] Initializing OpenEyes Vision Engine v3.0.1

[SYSTEM] Hardware detected: Jetson Orin Nano (8GB)

[SYSTEM] CUDA Available: True | TensorRT: True

[INFO] Loading YOLOv10n INT8 engine... DONE (1.2s)

[INFO] Initializing MediaPipe FaceMesh (max_faces=3)... DONE

[INFO] Initializing MediaPipe Hands... DONE

[INFO] Starting DeepStream pipeline via appsink...

FPS: 60 | Obj: 3 | Face: 1 | Hand: thumbs_up | Pose: 1

FPS: 60 | Obj: 4 | Face: 1 | Hand: open_palm | Pose: 1

FPS: 59 | Obj: 4 | Face: 1 | Hand: open_palm | Pose: 1

Core Architecture

⚡

DeepStream Pipeline

Hardware-accelerated processing via GStreamer/DeepStream. Runs TensorRT YOLO engines directly on GPU, passing frames via appsink for zero-copy Python manipulation.

👤

Multi-Modal Inference

Simultaneous FaceMesh (up to 3 faces), Hand tracking (8 defined gestures), and full-body Pose estimation running alongside primary object detection.

📡

ROS2 Native

Publishes telemetry across 10 specialized topics (/vision/detections, /vision/depth, /vision/pose) using MultiThreadedExecutor.

🧠

World Models

Predictive intelligence at 200 Hz with LeWM (15M params) and V-JEPA 2 for spatiotemporal awareness.

Performance Benchmarks

Tested on NVIDIA Jetson Orin Nano (8GB) in MAXN mode.

Configuration	Models Active	Frame Rate	Latency
Detection Only (INT8)	YOLOv10n TensorRT	60 FPS	16ms
Minimal Pipeline	Detection + Depth + Tracking	35-40 FPS	28ms
Full Pipeline (v3.0.1)	Detection + Face + Gesture + Pose	25-30 FPS	38ms
World Model Planning	LeWM 15M (Inference Only)	200 Hz	5ms

We Give RobotsVision