Deep Video Analytics    

by Akshay Bhat, Cornell University.

Deep Video Analytics is a platform for indexing and extracting information from videos and images. With latest version of docker installed correctly, you can run Deep Video Analytics in minutes locally (even without a GPU) using a single command.

For installation instructions & demo please visit https://www.deepvideoanalytics.com

Documentation & tutorial

• For a quick overview we strongly recommend going through the presentation in readme.pdf

• Documentation along with a tutorial is being written in /docs/tutorial directory.

Experiments

• OCR example has been moved to /docs/experiments/ocr directory.
• More experiments coming soon!

Deployment

We provide instructions for deploying DVA in three scenarios.

1. deploy/cpu contains docker-compose files for non-GPU single machine deployments on Linode, AWS, GCP etc.

2. deploy/gpu contains docker-compose files for GPU single machine deployments on GCP, AWS etc.

3. deploy/kube contains files used for launching DVA in a scalable GKE + GCS setup

Development

• deploy/dev contains docker-compose files for interactively developing DVA by using host server directory mapped as a volume.

Code organization

• /client : Python client using DVA REST API
• /configs : ngnix config + defaults.py defining models + processing pipelines (can be replaced by mounting a volume)
• /deploy : Dockerfiles + Instructions for development, single machine deployment abnd scalable deployment with Kubernetes
• /docs : Documentation, tutorial and experiments
• /tests : Files required for testing
• /repos : Code copied from third party repos, e.g. Yahoo LOPQ, TF-CTPN etc.
• /server : dvalib + django server contains contains bulk of the code for UI, App and models.
• /logs : Empty dir for storing logs

Libraries modified in code and their licenses

Library	Link to the license
YAD2K	MIT License
AdminLTE2	MIT License
FabricJS	MIT License
Facenet	MIT License
JSFeat	MIT License
MTCNN	MIT License
CRNN.pytorch	MIT License
Original CRNN code by Baoguang Shi	MIT License
Object Detector App using TF Object detection API	MIT License
Plotly.js	MIT License
CRF as RNN	MIT License
Text Detection CTPN	MIT License
SphereFace	MIT License
Segment annotator	BSD 3-clause
TF Object detection API	Apache 2.0
TF models/slim	Apache 2.0
TF models/delf	Apache 2.0
Youtube 8M feature extractor	Apache 2.0
CROW	Apache 2.0
LOPQ	Apache 2.0
Open Images Pre-trained network	Apache 2.0

Additional libraries & frameworks

• FFmpeg (not linked, called via a Subprocess)
• Tensorflow
• OpenCV
• Numpy
• Pytorch
• Docker
• Nvidia-docker
• Docker-compose
• All packages in requirements.txt
• All dependancies in Dockerfile

License & Copyright

Copyright 2016-2017, Akshay Bhat, Cornell University, All rights reserved.

Contact

Deep Video Analytics is currently in active development. The license will be relaxed once a stable release version is reached. Please contact me for more information. For more information see answer on this issue