- blink.cpp: eye-blink classification based on facial landmarks
- bow: BOWKmajority example(for binary descriptors) also, bow with python in opencv 3.x
- chili: opencv c++ remote playground running on heroku
- cluster: clustering att faces with kmeans,flann,random and BRUTEFORCE!!
- curl.cpp: download a car from the internet ;)
- colab: some jupyter notebooks, mostly tf/pytorch -> dnn conversions
- condens2.cpp: 2.x adaption of the old legacy condensation code
- db: serialize cv::Mat to some popular databases.
- dbrief: c++, opencv2.x port of http://cvlab.epfl.ch/research/detect/dbrief
- demo: various c++/py samples, mostly related to problems found on answers.opencv.org or #opencv
- dnn_dot: graphviz visualization of opencv's dnn models
- eyes.cpp: hog/svm based open / closed eyes detector, using flandmarks for eye location
- facerec_dnn: using opencv's dnn module and a pretrained (torch) facenet for transfer learning. also a realtime webcam demo.
- facerec_online.py: realtime facerec from webcam ( with adaptive face-db )
- facerec_online.cpp: same in c++
- gas.cpp: Neural Gas (clustering) experimentBRUTE FORCE!!!!
- gldemo.cpp : opengl inside opencv (+ minimal .obj reader)
- java_dnn: java examples for dnn models
- java_face3.0: testcase for ocv3 facerec bindings in java
- headpose: small head pose estimation (using dlib & pico)
- hep.cpp: playing with http://dismac.dii.unipg.it/hep/index.html
- html5: facedetect from webcam in browser ( using liu liu's js code )
- ica: An Introduction to Independent Component Analysis: InfoMax and FastICA algorithms
- kalman.cpp: mouse-tracking kalman demo
- kotlin: jikes, opencv java bindings in kotlin ;)
- mace: Minimum Average Correlation Energy Filter, useful for authentification with (cancellable) biometrical features from small datasets.
- mjpg_serve.py: mjpeg webcam server ( single-thread, single-client, sorry )
- mjpg_serve.cpp: mjpeg webcam server ( single-thread, but multi client )
- ml/libsvm_ocv.cpp : using libsvm with opencv
- openpose_multi.cpp: an attempt to use opencv's dnn with multiple persons, exploiting the PAF's in the original openpose model (WIP)
- profile.h: profiling / timing code
- shape_match: some shape/contours/gestures playground (various distances with a common interface).
- skin.cpp: skin detector (Rehg/Jones,1999)
- tcnn_landmarks: caffe_dnn, a small example for pretrained facial landmarks from tal hassner's tcnn project
- tsc: trafic sign classification with https://github.com/tiny-dnn/
- toy/echo: echo chamber
- toy/zoom: zooming in/out of highgui window
- toy/kaleidoscope: .. kids ..
- tracking: BACF and CMT tracker
- ui : onscreen ui (sliders & buttons & selectboxes)
- vibe.cpp: ViBe background separation
- violentFlow.cpp : http://www.openu.ac.il/home/hassner/data/violentflows/violent_flows.pdf