COCO_2018-Stuff-Segmentation-Challenge What is COCO? COCO is large-scale object detection, segmentation, and captioning dataset. The COCO Stuff Segmentation Task is designed to push the state of the art in semantic segmentation of stuff classes. Whereas the object detection task addresses thing classes (person, car, elephant), this task focuses on stuff classes (grass, wall, sky). For full details of the stuff segmentation task please see the stuff evaluation page. Note: the newly introduced panoptic segmentation task addresses recognition of both things and stuff classes simultaneously. This project implicates schematic segmentation, where raw RGB images are processed into the model and pixel masked images are given out as an output.

Internship project in Bennett University under Leadinginadi.ai

Problem Statement :-
To perform Semantic Segmentation of Stuff classes.The COCO Stuff Segmentation Task is designed to push the state of the art in semantic segmentation of stuff classes.

you can run script.sh in the scripts folder to install all prequisites. selective prequisites are given below. pip install tensorflow-gpu
pip install tqdm
pip install keras
pip install keras-segmentation

You need to make two folders

Images Folder - For all the training images
Annotations Folder - For the corresponding ground truth segmentation images
The filenames of the annotation images should be same as the filenames of the RGB images.

python -m keras_segmentation verify_dataset \
 --images_path="dataset1/images_prepped_train/" \
 --segs_path="dataset1/annotations_prepped_train/"  \
 --n_classes=50

python -m keras_segmentation visualize_dataset \
 --images_path="dataset1/images_prepped_train/" \
 --segs_path="dataset1/annotations_prepped_train/"  \
 --n_classes=50

python -m keras_segmentation train \
--checkpoints_path="path_to_checkpoints" \
--train_images="dataset1/images_prepped_train/" \
--train_annotations="dataset1/annotations_prepped_train/" \
--val_images="dataset1/images_prepped_test/" \
--val_annotations="dataset1/annotations_prepped_test/" \
--n_classes=300 \
--input_height=320 \
--input_width=640 \
--model_name="pspnet"

python -m keras_segmentation predict \
 --checkpoints_path="path_to_checkpoints" \
 --input_path="dataset1/images_prepped_test/" \
 --output_path="path_to_predictions"

1. https://github.com/divamgupta/image-segmentation-keras
   
2. https://github.com/GeorgeSeif/Semantic-Segmentation-Suite
   
3. https://github.com/aurora95/Keras-FCN
   
4. https://cocodataset.org/#stuff-2018
   
5. https://arxiv.org/abs/1612.03716