Skip to content

Code release for paper "A Modulation Module for Multi-task Learning with Applications in Image Retrieval"

License

Notifications You must be signed in to change notification settings

Zhaoxiangyun/Multi-Task-Modulation-Module

Repository files navigation

A Modulation Module for Multi-task Learning with Applications in Image Retrieval

This is a TensorFlow implementation of our ECCV paper "A Modulation Module for Multi-task Learning with Applications in Image Retrieval". The released code include the training and testing code for 7 attributes. This code could support training and testing with any number of training attributes. The code can be easily changed for 20 and 40 number of attributes training. We have released the prepared training data and testing data for 7 attributes. The training data supports training with any number of attributes up to 40 attributes in the dataset. The testing data for other attributes will come soon.

Requirement

The code is tested using Tensorflow 1.0.0 under Ubuntu 14.04 with Python 3.5. The code is based on the facenet impelementation. You could refer it for the software requirements.

Training and testing data

The CelebA dataset has been used for training and testing. This training set consists of around ranmdomly selected 30,000 image pairs over 40 attributes after face detection. We have prepared the training here and testing data here.

Running training and testing

./experiments/train_mm.sh ./experiments/test_mm.sh

Performance

You could find the trained model [here]

About

Code release for paper "A Modulation Module for Multi-task Learning with Applications in Image Retrieval"

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published