Blog
Practical notes on machine learning, computer vision, Python, and the details worth understanding.
2026
-
MediaPipe Face Mesh in ONNX: 468 Dense 3D Landmarks
MediaPipe Face Mesh on ONNX Runtime: 468 dense 3D landmarks from a 192x192 crop, with BlazeFace bundled and parity checks against mp.solutions. -
FaceAttribNet: Eye, Glasses, and Mask Detection in ONNX
FaceAttribNet predicts eye openness, eyeglasses, sunglasses, and mask attributes from a 128x128 face crop using five independent binary heads. -
Python Core Interview Questions and Answers: Clear, Practical Explanations
Practical Python interview answers for core topics like data structures, object identity, memory management, the GIL, generators, decorators, and async FastAPI behavior.
2025
-
UniFace: A Unified Face Analysis Library for Python
UniFace is a Python library for face analysis on ONNX Runtime: six detectors, five recognition models, 468-point mesh, parsing, attributes, and quality scoring.
2024
-
Face Parsing with BiSeNet and ResNet Backbones
Face parsing with BiSeNet, ResNet18 and ResNet34 backbones, CelebAMask-HQ training, PyTorch inference, and ONNX export. -
Tiny-Face: Ultra-Lightweight Face Detection for Edge Devices
Tiny-Face compares SlimFace, RFB, and compact RetinaFace variants for small face detection models with WIDER FACE results and PyTorch/ONNX weights. -
RetinaFace: Single-Stage Face Detection in PyTorch
A RetinaFace implementation with MobileNet and ResNet backbones, WIDER FACE evaluation, webcam inference, ONNX support, and pretrained weights. -
The Geometric Perspective of Vectors in Machine Learning
Why geometric intuition for vectors and matrix transformations matters in machine learning, with practical examples and recommended learning resources. -
MobileGaze: Lightweight Gaze Estimation with MobileOne
MobileGaze estimates gaze direction with ResNet, MobileNet, and MobileOne backbones. Gaze360-trained, with PyTorch and ONNX weights for real-time inference. -
Real-Time Head Pose Estimation with MobileNet and ResNet
Head pose estimation with ResNet and MobileNet backbones, SCRFD face detection, PyTorch and ONNX weights, and pitch/yaw/roll output for real-time video.