Expertise
“What we cannot create, we do not understand.” - Richard Feynman
Diesimo Lab - My Edge AI Homelab
Solution Map
From research to production-ready real-world Edge AI and Computer Vision solutions for smart industries.
Edge AI
├── End-to-End Edge AI Stack
│ ├── Data Acquisition (Sensors: Camera, LiDAR ...)
│ ├── Model Design (Efficiency-Aware, NAS ...)
│ ├── Optimization (Quantization, Prunning, Distillation, fine-tuning )
│ ├── Deployment (ONNX/Onnxruntime, LiteRT, AIMET, TensorRT, OpenVINO, ...)
│ └── Profiling/Monitoring (Qualcomm AI Hub, Edge Impulse..)
├── Hardware-Aware Optimization
│ ├── Runtime/inference engine: Onnxruntime, custom inference engines
│ ├── Energy/Power Consumption (OPS/W): Battery-based devices (W => mW)
│ └── AI Accelerators: Multi-CPUs, GPUs, NPUs
└── Deployment at the Edge: Target Platform
├── Smart Devices: Smartphones, Wearable Devices
└── Embedded/IoT Devices: MCUs, Raspberry Pi, NVIDIA Jetson Nano
Computer Vision
├── End-to-End Computer Vision Stack
│ ├── Data Acquisition/Labeling
│ ├── Model Design (Efficiency-aware)
│ ├── Deployment (Edge Devices, Cloud)
│ └── Monitoring
└── Computer Vision Applications
├── Classification
├── Object Detection (2D, 3D)
├── Segmentation (semantic, instance, panoptic)
├── Multimodality (VLMs: GANs, VAEs, Diffusers ...)
└── Motion & Video Analysis: Tracking & Flow
Perception
└── Perception
├── Camera Modeling & Calibration (Monocular, Stereo, Fisheye)
├── Sensor Fusion (Camera, LiDAR, RADAR ...)
├── Depth Perception
└── VSLAM (VO)
Industry: Case Studies
├── Autonomous Systems
│ ├── Robotics # Mobile, Humanoids, Embodied Agents
│ ├── Self-Driving Cars # ADAS, Autonomous Vehicles
│ └── UAS: Drones, sUAS, UAVs # Unmanned Aerial Systems
├── Healthcare
│ └── Medical Imaging
├── Agriculture
│ └── Precision Farming
├── Security
│ └── Surveillance
├── Manufacturing
│ └── Quality Control
├── Retail
│ └── Consumer Analytics
└── Smart Cities
└── Urban Planning
Demos: POCS and MVS …
└── Demos: POCs & Minimum Viable Solution (MVS)
├── Colab, Jupyter Notebooks
├── Hugging Face Spaces: Gradio, Streamlit, Docker
├── Frontend Web/App UI: V0, Qt
└── Realtime App/Model Serving: Flask, FastAPI, WebSocket/WSS