aninggo / deepnet-for-semantic-labeling-photogrammetry

2D semantic segmentation that assigns labels to multiple object categories on Potsdam dataset :rocket:

DeepNet-for-Semantic-Labeling-Photogrammetry

This deep learning project is my graduated thesis which is scored 9.5/10. It is also the first time i do end-to-end an applied research project by myself.

This project aim to automated extraction of urban objects from data acquired by airborne sensors which very high-resolution data (leads to high intra-class variance while the inter-class variance is low), focus is on detailed 2D semantic segmentation that assigns labels to multiple object categories. At the end, my accuraccy reachs 90% (the top world leader at 92%)

Dataset:

• 2D Semantic Labeling Contest - Potsdam
• Short description at here

Pipeline

1. Visualise and analyse data
2. Preprocessing data
3. Build model
4. Choose hyper-parameters and loss function
5. Experiment and validate
6. Testing phase
7. Report and submit

1. Visualise and analyse data

See the detail. Conclusion:

• Unbalance dataset: building, road and low vegetable class have highest acreage, while car class is negligible.
• Large file size, high resolution.
• The tree class looks like low veg. class on the RGB images. So we use additional DSM channel to distinguish.
• The clutter class 's (disregard class, not the main class) percentage is quite big.

2. Preprocessing data

• Image size 6000x6000 5 channel (IR-R-G-DSM-nDSM), split to patchs 256x256
• Train/Valid/Test: 18/6/14

3 & 4. Build model and Experiment

I tried lots of models, but there are 3 main:

My Models	Origin Paper	My customize code	Acc
Fully Convolutional Networks with Resnet 101 encoder, 3 skip connections	FCN paper	My customize implement	0.85
Unet with Resnet 101 encoder	Unet paper	My customize implement	0.89
Pyramid Spatial Pooling Network with Resnet 101 encoder, aulixiary loss	PSPNet paper	My customize implement	0.90
90% (aulixiary loss) (single model)

4. Choose hyper-parameters and loss function

5. Experiment and validate

6. Testing phase

7. Report and submit

Expriment

More

• Use cyclical learning rate
• Use Resnet101 pretrain
• Split to maxable size patch
• Choose maxable batch size (6->32)
• Adam optimizer
• Small weight decay lr
• Use less pooling
• Use more skip connection
• Normalize image before training

Guide to run

Install package and run notebook on your computer (with anaconda, or cloud such as kaggle, colab)