/TextBoxes-TensorFlow

TextBoxes re-implement using tensorflow

Primary LanguageJupyter Notebook

TextBoxes-TensorFlow

TextBoxes re-implementation using tensorflow. This project is greatly inspired by slim project And many functions are modified based on SSD-tensorflow project Later, we will overwrite this project so make it more flexiable and modularized.

Author: Daitao Xing : dx383@nyu.edu Jin Huang : jh5442@nyu.edu

Progress

2017/ 03/14  

data_processing phase finished Test：

1. Download the dataset， put 1/ folder and gt.mat uner ddata/sythtext/ folder（will wirte script）   
2. python datasets/data2record.py    
3. python image_processing.py

output： batch_size * 300 * 300 * 3 image

2017/ 03/17  

Finish the design of training(can start training)

TASET_DIR=./data/sythtext/
TRAIN_DIR=./logs/
python Train.py \
--train_dir=${TRAIN_DIR} \
	--dataset_dir=${DATASET_DIR} \
	--save_summaries_secs=60 \
	--save_interval_secs=600 \
	--weight_decay=0.0005 \
	--optimizer=adam \
	--learning_rate=0.001 \
	--batch_size=16 \
	--num_clones=1 \
	--gpu_memory_fraction=0.5

Problems to be solved：

1. Need to redesign visualization		
2. image_processing can be improved

Next steps:

  1. traing on other datasets
  2. fine tunes
  3. test
  4. automatic downloading datasets and so on