The main N2D2 binary, which allows to run DNN learning, testing, benchmarking and export.
Just run ./n2d2 -h to get the full list of program options.
A shell helper for launching n2d2 processes in sub-directory.
This binary allows you to run a classification DNN live from a webcam or a
video.
It works with 1D output layer (generally softmax or fully-connected) networks
equiped with a Target object. See the application examples for a
use-case.
This binary allows you to run a segmentation and classification DNN of type
"fully-CNN" live from a webcam or a video. It works with 2D output layer
networks equiped with a TargetROIs object. See the application examples for
a use-case.
The following application examples are provided:
A live face detection application, with gender recognition, based on the IMDB-WIKI dataset. You will need a webcam supported by OpenCV to run this application live, or you can run it on a video file.
A live object recognition application, based on ILSVRC2012 (ImageNet) dataset. You will need a webcam supported by OpenCV to run this application live, or you can run it on a video file.
A road segmentation application, based on the KITTI Road dataset.
This simulation implements STDP unsupervised learning on a recorded AER sequence of cars running on a highway, using the event-based simulator embedded into N2D2.
This binary reproduces some of the results published in @Bichler2011.
This binary is a simple DVS128 format AER viewer.
[@Bichler2011]: O. Bichler, D. Querlioz, S. Thorpe, J. Bourgoin, and C. Gamrat. Extraction of temporally correlated features from dynamic vision sensors with spike-timing-dependent plasticity. Neural Networks, 32:339-348, 2012. doi:[10.1016/j.neunet.2012.02.022] (http://dx.doi.org/10.1016/j.neunet.2012.02.022).