Merge pull request adipandas#12 from adipandas/devel

Merge development branch

Author: adipandas

.gitignore

@@ -1,10 +1,10 @@
 .ipynb_checkpoints/
-video_data/
+examples/video_data/
 .idea/
 examples/output.avi
-pretrained_models/caffemodel_weights/
-pretrained_models/tensorflow_weights/
-pretrained_models/yolo_weights/
+examples/pretrained_models/caffemodel_weights/
+examples/pretrained_models/tensorflow_weights/
+examples/pretrained_models/yolo_weights/
 
 # Byte-compiled / optimized / DLL files
 __pycache__/

DOWNLOAD_WEIGHTS.md

@@ -0,0 +1,22 @@
+## Instructions to download pretrained neural-network weights.
+
+##### YOLOv3
+```
+cd ./examples/pretrained_models/yolo_weights
+sudo chmod +x ./get_yolo.sh
+./get_yolo.sh
+```
+
+##### TensorFlow - MobileNetSSDv2
+```
+cd ./pretrained_models/tensorflow_weights
+sudo chmod +x ./get_ssd_model.sh
+./get_ssd_model.sh
+```
+
+##### Caffemodel - MobileNetSSD
+```
+cd ./pretrained_models/caffemodel_weights
+sudo chmod +x ./get_caffemodel.sh
+./get_caffemodel.sh
+```

README.md

@@ -1,30 +1,37 @@
-[cars-yolo-output]: ./assets/cars.gif "Sample Output with YOLO"
-[cows-tf-ssd-output]: ./assets/cows.gif "Sample Output with SSD"
+[cars-yolo-output]: examples/assets/cars.gif "Sample Output with YOLO"
+[cows-tf-ssd-output]: examples/assets/cows.gif "Sample Output with SSD"
 
 # Multi-object trackers in Python
-Object detection using deep learning and multi-object tracking
+Various multi-object tracking algorithms.
 
 [![DOI](https://zenodo.org/badge/148338463.svg)](https://zenodo.org/badge/latestdoi/148338463)
 
-### Available Trackers
-```
-SimpleTracker
 
-SimpleTracker2
-```
+`YOLOv3 + CentroidTracker` |  `TF-MobileNetSSD + CentroidTracker`
+:-------------------------:|:-------------------------:
+![Cars with YOLO][cars-yolo-output]  |  ![Cows with tf-SSD][cows-tf-ssd-output]
+Video source: [link](https://flic.kr/p/L6qyxj) | Video source: [link](https://flic.kr/p/26WeEWy)
 
-#### YOLO
-Video Source: [link](https://flic.kr/p/89KYXt)
 
-![Cars with YOLO][cars-yolo-output]
+## Available Multi Object Trackers
 
-#### Tensorflow-SSD-MobileNet
-Video Source: [link](https://flic.kr/p/26WeEWy)
+```
+CentroidTracker
+IOUTracker
+CentroidKF_Tracker
+SORT
+```
 
-![Cows with tf-SSD][cows-tf-ssd-output]
+## Available OpenCV-based object detectors:
 
+```
+detector.TF_SSDMobileNetV2
+detector.Caffe_SSDMobileNet
+detector.YOLOv3
+```
+
+## Installation
 
-### Installation
 Pip install for OpenCV (version 3.4.3 or later) is available [here](https://pypi.org/project/opencv-python/) and can be done with the following command:
 
 ```
@@ -43,56 +50,30 @@ cd multi-object-tracker
 pip install -e .
 ```
 
-### YOLO
-
-Do the following in the terminal to download a pretrained weights of YOLO:
-```
-cd ./pretrained_models/yolo_weights
-sudo chmod +x ./get_yolo.sh
-./get_yolo.sh
-```
-
-### TensorFlow model
-
-Do the following in the terminal to download a pretrained model:
-```
-cd ./pretrained_models/tensorflow_weights
-sudo chmod +x ./get_ssd_model.sh
-./get_ssd_model.sh
-```
+Note - for using neural network models with GPU
+---
+For using the opencv `dnn`-based object detection modules provided in this repository with GPU, you may have to compile a CUDA enabled version of OpenCV from source.
 
-**SSD-Mobilenet_v2_coco_2018_03_29** was used for this example.
-Other networks can be downloaded and ran: Go through `tracking-tensorflow-ssd_mobilenet_v2_coco_2018_03_29.ipynb` for more details.
-
-### Caffemodel
-
-Do the following in the terminal to download a pretrained model:
-```
-cd ./pretrained_models/caffemodel_weights
-sudo chmod +x ./get_caffemodel.sh
-./get_caffemodel.sh
-```
+For building opencv from source, you can refer the following:
+[[link-1](https://docs.opencv.org/master/df/d65/tutorial_table_of_content_introduction.html)],
+[[link-2](https://www.pyimagesearch.com/2020/02/03/how-to-use-opencvs-dnn-module-with-nvidia-gpus-cuda-and-cudnn/)]
 
-This is a MobileNet-SSD caffemodel.
+## How to use?: Examples
 
-### Examples and How to use:
+Please refer [examples](./examples/) folder of this repository.
+You can clone and run the examples as shown in the [readme](examples/readme.md) inside the [examples](./examples/) folder.
 
-For examples and how to use this repository, please refer [examples/](examples/) folder.
+## Pretrained object detection models
 
-### References and Credits
-This work is based on the following literature:
-1. Bochinski, E., Eiselein, V., & Sikora, T. (2017, August). High-speed tracking-by-detection without using image information. In 2017 14th IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS) (pp. 1-6). IEEE. [[paper-pdf](http://elvera.nue.tu-berlin.de/files/1517Bochinski2017.pdf)]
-2. Pyimagesearch [link-1](https://www.pyimagesearch.com/2018/07/23/simple-object-tracking-with-opencv/), [link-2](https://www.pyimagesearch.com/2018/11/12/yolo-object-detection-with-opencv/)
-3. [correlationTracker](https://github.com/Wenuka/correlationTracker)
-4. [Caffemodel zoo](http://caffe.berkeleyvision.org/model_zoo.html)
-5. [Caffemodel zoo GitHub](https://github.com/BVLC/caffe/tree/master/models)
-6. [YOLO v3](https://pjreddie.com/media/files/papers/YOLOv3.pdf)
+You will have to download the pretrained weights for the neural-network models. 
+The shell scripts for downloading these are provided in [examples](examples/) folder.  
+Please refer [DOWNLOAD_WEIGHTS.md](DOWNLOAD_WEIGHTS.md) for more details.
 
-Use the caffemodel zoo from the reference [4,5] mentioned above to vary the CNN models and Play around with the codes.
+## References and Credits
 
-***Suggestion**: If you are looking for speed go for SSD-mobilenet. If you are looking for accurracy and speed go with YOLO. The best way is to train and fine tune your models on your dataset. Although, Faster-RCNN gives more accurate object detections, you will have to compromise on the detection speed as it is slower as compared to YOLO.*
+Please see [REFERENCES.md](REFERENCES.md)
 
-### Citation
+## Citation
 
 If you use this repository in your work, please consider citing it with:
 ```

REFERENCES.md

@@ -0,0 +1,13 @@
+# References and Credits
+
+This work is based on the following literature:
+
+1. Bochinski, E., Eiselein, V., & Sikora, T. (2017, August). High-speed tracking-by-detection without using image information. In 2017 14th IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS) (pp. 1-6). IEEE. [[pdf](http://elvera.nue.tu-berlin.de/files/1517Bochinski2017.pdf)]
+2. Bewley, A., Ge, Z., Ott, L., Ramos, F., & Upcroft, B. (2016, September). Simple online and realtime tracking. In 2016 IEEE International Conference on Image Processing (ICIP) (pp. 3464-3468). IEEE. [[arxiv](https://arxiv.org/abs/1602.00763)]
+3. YOLOv3. [[pdf](https://pjreddie.com/media/files/papers/YOLOv3.pdf)][[website](https://pjreddie.com/darknet/yolo/)]
+4. Kalman Filter. [[wiki](https://en.wikipedia.org/wiki/Kalman_filter)]
+5. TensorFlow Object Detection API [[github](https://github.com/tensorflow/models/tree/master/research/object_detection)]
+6. Caffe [[website](https://caffe.berkeleyvision.org/)][[github](https://github.com/BVLC/caffe)]
+
+
+#### Link to `multi-object-tracker` [[webpage](https://adipandas.github.io/multi-object-tracker/)][[GitHub Repo](https://github.com/adipandas/multi-object-tracker)]

_config.yml

@@ -1,2 +1,3 @@
 title: Multi-object trackers in Python
+description:
 theme: jekyll-theme-cayman