Home

Awesome

Egocentric Video Description based on Temporally-Linked Sequences

This repository contains the code for building the Temporally-linked Multi-input Attention (TMA) model, which was presented in the work Egocentric Video Description based on Temporally-Linked Sequences, submitted to the Journal of Visual Communication and Image Representation. With this module, you can replicate our experiments and easily deploy new models. TMA is built upon our fork of Keras framework (version 1.2) and tested for the Theano backend.

Features:

Architecture

TMA_model

Requirements

TMA requires the following libraries:

Instructions:

Assuming you have a dataset and features extracted from the video frames:

  1. Set the paths to Keras and Multimodal Keras Wraper in train.sh

  2. Prepare data:

python data_engine/subsample_frames_features.py

python data_engine/generate_features_lists.py

python data_engine/generate_descriptions_lists.py

See data_engine/README.md for detailed information.

  1. Prepare the inputs/outputs of your model in data_engine/prepare_data.py

  2. Set a model configuration in config.py

  3. Train!:

python main.py

Dataset

The dataset EDUB-SegDesc was used to evaluate this model. It was acquired by the wearable camera Narrative Clip, taking a picture every 30 seconds (2 fpm). It consists of 55 days acquired by 9 people. Containing a total of 48,717 images, divided in 1,339 events (or image sequences) and 3,991 captions.

Citation

If you use this code for any purpose, please, do not forget to cite the following paper:

Marc Bolaños, Álvaro Peris, Francisco Casacuberta, Sergi Soler and Petia Radeva.
Egocentric Video Description based on Temporally-Linked Sequences
In Special Issue on Egocentric Vision and Lifelogging Tools. 
Journal of Visual Communication and Image Representation (VCIR), (SUBMITTED).

About

Joint collaboration between the Computer Vision at the University of Barcelona (CVUB) group at Universitat de Barcelona-CVC and the PRHLT Research Center at Universitat Politècnica de València.

Contact

Marc Bolaños (web page): marc.bolanos@ub.edu

Álvaro Peris (web page): lvapeab@prhlt.upv.es