ronghanghu / Text_objseg
Licence: other
Code release for Hu et al. Segmentation from Natural Language Expressions. in ECCV, 2016
Stars: ✭ 86
Labels
Projects that are alternatives of or similar to Text objseg
Training Material
A collection of code examples as well as presentations for training purposes
Stars: ✭ 85 (-1.16%)
Mutual labels: jupyter-notebook
Wotan
Automagically remove trends from time-series data
Stars: ✭ 86 (+0%)
Mutual labels: jupyter-notebook
Quantum programming tutorial
Gamified tutorial for the QISKit quantum SDK
Stars: ✭ 86 (+0%)
Mutual labels: jupyter-notebook
Introdatasci
Course materials for: Introduction to Data Science and Programming
Stars: ✭ 86 (+0%)
Mutual labels: jupyter-notebook
Network science meets deep learning
Designing Deep neural network architectures using topologies from the world of Complex Networks/network Science
Stars: ✭ 86 (+0%)
Mutual labels: jupyter-notebook
Kaggle Competitions
There are plenty of courses and tutorials that can help you learn machine learning from scratch but here in GitHub, I want to solve some Kaggle competitions as a comprehensive workflow with python packages. After reading, you can use this workflow to solve other real problems and use it as a template.
Stars: ✭ 86 (+0%)
Mutual labels: jupyter-notebook
100 Plus Python Programming Exercises Extended
100+ python programming exercise problem discussed ,explained and solved in different ways
Stars: ✭ 1,250 (+1353.49%)
Mutual labels: jupyter-notebook
Deep Learning Boot Camp
A community run, 5-day PyTorch Deep Learning Bootcamp
Stars: ✭ 1,270 (+1376.74%)
Mutual labels: jupyter-notebook
Sagemaker Ml Workflow With Apache Airflow
This repository shows a sample example to build, manage and orchestrate Machine Learning workflows using Amazon Sagemaker and Apache Airflow.
Stars: ✭ 86 (+0%)
Mutual labels: jupyter-notebook
Breast Cancer Classification
Breast Cancer Classification using CNN and transfer learning
Stars: ✭ 86 (+0%)
Mutual labels: jupyter-notebook
Book Mlearn Gyomu
Book sample (AI Machine-learning Deep-learning)
Stars: ✭ 84 (-2.33%)
Mutual labels: jupyter-notebook
Sphinx Book Theme
A lightweight book theme built off of the pydata sphinx theme
Stars: ✭ 86 (+0%)
Mutual labels: jupyter-notebook
Viz torch optim
Videos of deep learning optimizers moving on 3D problem-landscapes
Stars: ✭ 86 (+0%)
Mutual labels: jupyter-notebook
Caffeonspark
Distributed deep learning on Hadoop and Spark clusters.
Stars: ✭ 1,272 (+1379.07%)
Mutual labels: jupyter-notebook
Python For Data Scientists
Deliverable: This Jupyter notebook will help aspiring data scientists learn and practice the necessary python code needed for many data science projects.
Stars: ✭ 86 (+0%)
Mutual labels: jupyter-notebook
Segmentation from Natural Language Expressions
This repository contains the code for the following paper:
- R. Hu, M. Rohrbach, T. Darrell, Segmentation from Natural Language Expressions. in ECCV, 2016. (PDF)
@article{hu2016segmentation,
title={Segmentation from Natural Language Expressions},
author={Hu, Ronghang and Rohrbach, Marcus and Darrell, Trevor},
journal={Proceedings of the European Conference on Computer Vision (ECCV)},
year={2016}
}
Project Page: http://ronghanghu.com/text_objseg
Installation
- Install Google TensorFlow (v1.0.0 or higher) following the instructions here.
- Download this repository or clone with Git, and then
cd
into the root directory of the repository.
Demo
- Download the trained models:
exp-referit/tfmodel/download_trained_models.sh
. - Run the language-based segmentation model demo in
./demo/text_objseg_demo.ipynb
with Jupyter Notebook (IPython Notebook).
Training and evaluation on ReferIt Dataset
Download dataset and VGG network
- Download ReferIt dataset:
exp-referit/referit-dataset/download_referit_dataset.sh
. - Download VGG-16 network parameters trained on ImageNET 1000 classes:
models/convert_caffemodel/params/download_vgg_params.sh
.
Training
- You may need to add the repository root directory to Python's module path:
export PYTHONPATH=.:$PYTHONPATH
. - Build training batches for bounding boxes:
python exp-referit/build_training_batches_det.py
. - Build training batches for segmentation:
python exp-referit/build_training_batches_seg.py
. - Select the GPU you want to use during training:
export GPU_ID=<gpu id>
. Use 0 for<gpu id>
if you only have one GPU on your machine. - Train the language-based bounding box localization model:
python exp-referit/exp_train_referit_det.py $GPU_ID
. - Train the low resolution language-based segmentation model (from the previous bounding box localization model):
python exp-referit/init_referit_seg_lowres_from_det.py && python exp-referit/exp_train_referit_seg_lowres.py $GPU_ID
. - Train the high resolution language-based segmentation model (from the previous low resolution segmentation model):
python exp-referit/init_referit_seg_highres_from_lowres.py && python exp-referit/exp_train_referit_seg_highres.py $GPU_ID
.
Alternatively, you may skip the training procedure and download the trained models directly:
exp-referit/tfmodel/download_trained_models.sh
.
Evaluation
- Select the GPU you want to use during testing:
export GPU_ID=<gpu id>
. Use 0 for<gpu id>
if you only have one GPU on your machine. Also, you may need to add the repository root directory to Python's module path:export PYTHONPATH=.:$PYTHONPATH
. - Run evaluation for the high resolution language-based segmentation model:
python exp-referit/exp_test_referit_seg.py $GPU_ID
This should reproduce the results in the paper. - You may also evaluate the language-based bounding box localization model:
python exp-referit/exp_test_referit_det.py $GPU_ID
The results can be compared to this paper.
Note that the project description data, including the texts, logos, images, and/or trademarks,
for each open source project belongs to its rightful owner.
If you wish to add or remove any projects, please contact us at [email protected].