GitHub - gdiprisco/Multitask_CNNs_for_efficient_face_analysis_in_the_wild: An innovative multi-task neural architecture suitable for face analysis and the consequent identification of gender, age, ethnicity and emotion. A network capable of the performance of the state of the art of single-task networks by sharing the feature extraction part. It may be suitable for robotic applications: it requires time and memory in the order of a single-task network.

Reference SE-ResNet-50 backbones implementation: https://github.com/qubvel/classification_models

Name		Name	Last commit message	Last commit date
Latest commit History 18 Commits
dataset		dataset
evaluate		evaluate
training		training
README.md		README.md
content.txt		content.txt
requirements.txt		requirements.txt

Provide feedback