Crowd Counting using Xception and CSRNet

This is our final project for the Vision and Cognitive Systems course at the University of Padua.

Task: Crowd Counting

Approaches:

Regression-based: Xception
- MAE: 1.5996
Density-based: CSRNet
- MAE: 3.44 (with pre-training) and 2.72 (CSRNet as a feature extractor)

Group members:

Name		Name	Last commit message	Last commit date
Latest commit History 7 Commits
1_Regression_Baseline_Xception_Training.ipynb		1_Regression_Baseline_Xception_Training.ipynb
2_Regression_Baseline_Model_Comparison.ipynb		2_Regression_Baseline_Model_Comparison.ipynb
3_Xception_with_pretraining_procedure.ipynb		3_Xception_with_pretraining_procedure.ipynb
4_CSRNet_FromScratch.ipynb		4_CSRNet_FromScratch.ipynb
5_CSRNet_TransferLearning_FineTunning_and_Regressor.ipynb		5_CSRNet_TransferLearning_FineTunning_and_Regressor.ipynb
Presentation.pdf		Presentation.pdf
README.md		README.md
Report.pdf		Report.pdf

Provide feedback