grasp_det.py

#!/usr/local/bin/python
''' 
Training a network on cornell grasping dataset for detecting grasping positions.
'''
import sys
import argparse
import os.path
import glob
import tensorflow as tf
import numpy as np
from shapely.geometry import Polygon
import grasp_img_proc
from grasp_inf import inference
import time

TRAIN_FILE = '/root/dataset/cornell_grasping_dataset/train-cgd'
VALIDATE_FILE = '/root/dataset/cornell_grasping_dataset/validation-cgd'

def bboxes_to_grasps(bboxes):
    # converting and scaling bounding boxes into grasps, g = {x, y, tan, h, w}
    box = tf.unstack(bboxes, axis=1)
    x = (box[0] + (box[4] - box[0])/2) * 0.35
    y = (box[1] + (box[5] - box[1])/2) * 0.47
    tan = (box[3] -box[1]) / (box[2] -box[0]) *0.47/0.35
    h = tf.sqrt(tf.pow((box[2] -box[0])*0.35, 2) + tf.pow((box[3] -box[1])*0.47, 2))
    w = tf.sqrt(tf.pow((box[6] -box[0])*0.35, 2) + tf.pow((box[7] -box[1])*0.47, 2))
    return x, y, tan, h, w

def grasp_to_bbox(x, y, tan, h, w):
    theta = tf.atan(tan)
    edge1 = (x -w/2*tf.cos(theta) +h/2*tf.sin(theta), y -w/2*tf.sin(theta) -h/2*tf.cos(theta))
    edge2 = (x +w/2*tf.cos(theta) +h/2*tf.sin(theta), y +w/2*tf.sin(theta) -h/2*tf.cos(theta))
    edge3 = (x +w/2*tf.cos(theta) -h/2*tf.sin(theta), y +w/2*tf.sin(theta) +h/2*tf.cos(theta))
    edge4 = (x -w/2*tf.cos(theta) -h/2*tf.sin(theta), y -w/2*tf.sin(theta) +h/2*tf.cos(theta))
    return [edge1, edge2, edge3, edge4]

def run_training():
    print(FLAGS.train_or_validation)
    if FLAGS.train_or_validation == 'train':
        print('distorted_inputs')
        data_files_ = TRAIN_FILE
        images, bboxes = grasp_img_proc.distorted_inputs(
                  [data_files_], FLAGS.num_epochs, batch_size=FLAGS.batch_size)
    else:
        print('inputs')
        data_files_ = VALIDATE_FILE
        images, bboxes = grasp_img_proc.inputs([data_files_])
    
    x, y, tan, h, w = bboxes_to_grasps(bboxes)
    x_hat, y_hat, tan_hat, h_hat, w_hat = tf.unstack(inference(images), axis=1) # list
    # tangent of 85 degree is 11 
    tan_hat_confined = tf.minimum(11., tf.maximum(-11., tan_hat))
    tan_confined = tf.minimum(11., tf.maximum(-11., tan))
    # Loss function
    gamma = tf.constant(10.)
    loss = tf.reduce_sum(tf.pow(x_hat -x, 2) +tf.pow(y_hat -y, 2) + gamma*tf.pow(tan_hat_confined - tan_confined, 2) +tf.pow(h_hat -h, 2) +tf.pow(w_hat -w, 2))
    train_op = tf.train.AdamOptimizer(epsilon=0.1).minimize(loss)
    init_op = tf.group(tf.global_variables_initializer(), tf.local_variables_initializer())
    sess = tf.Session()
    sess.run(init_op)
    coord = tf.train.Coordinator()
    threads = tf.train.start_queue_runners(sess=sess, coord=coord)
    #save/restore model
    d={}
    l = ['w1', 'b1', 'w2', 'b2', 'w3', 'b3', 'w4', 'b4', 'w5', 'b5', 'w_fc1', 'b_fc1', 'w_fc2', 'b_fc2']
    for i in l:
        d[i] = [v for v in tf.get_collection(tf.GraphKeys.GLOBAL_VARIABLES) if v.name == i+':0'][0]
    
    dg={}
    lg = ['w1', 'b1', 'w2', 'b2', 'w3', 'b3', 'w4', 'b4', 'w5', 'b5', 'w_fc1', 'b_fc1', 'w_fc2', 'b_fc2', 'w_output', 'b_output']
    for i in lg:
        dg[i] = [v for v in tf.get_collection(tf.GraphKeys.GLOBAL_VARIABLES) if v.name == i+':0'][0]

    saver = tf.train.Saver(d)
    saver_g = tf.train.Saver(dg)
    #saver.restore(sess, "/root/grasp/grasp-detection/models/imagenet/m2/m2.ckpt")
    saver_g.restore(sess, FLAGS.model_path)
    try:
        count = 0
        step = 0
        start_time = time.time()
        while not coord.should_stop():
            start_batch = time.time()
            #train
            if FLAGS.train_or_validation == 'train':
                _, loss_value, x_value, x_model, tan_value, tan_model, h_value, h_model, w_value, w_model = sess.run([train_op, loss, x, x_hat, tan, tan_hat, h, h_hat, w, w_hat])
                duration = time.time() - start_batch
                if step % 100 == 0:             
                    print('Step %d | loss = %s\n | x = %s\n | x_hat = %s\n | tan = %s\n | tan_hat = %s\n | h = %s\n | h_hat = %s\n | w = %s\n | w_hat = %s\n | (%.3f sec/batch\n')%(step, loss_value, x_value[:3], x_model[:3], tan_value[:3], tan_model[:3], h_value[:3], h_model[:3], w_value[:3], w_model[:3], duration)
                if step % 1000 == 0:
                    saver_g.save(sess, FLAGS.model_path)
            else:
                bbox_hat = grasp_to_bbox(x_hat, y_hat, tan_hat, h_hat, w_hat)
                bbox_value, bbox_model, tan_value, tan_model = sess.run([bboxes, bbox_hat, tan, tan_hat])
                bbox_value = np.reshape(bbox_value, -1)
                bbox_value = [(bbox_value[0]*0.35,bbox_value[1]*0.47),(bbox_value[2]*0.35,bbox_value[3]*0.47),(bbox_value[4]*0.35,bbox_value[5]*0.47),(bbox_value[6]*0.35,bbox_value[7]*0.47)] 
                p1 = Polygon(bbox_value)
                p2 = Polygon(bbox_model)
                iou = p1.intersection(p2).area / (p1.area +p2.area -p1.intersection(p2).area) 
                angle_diff = np.abs(np.arctan(tan_model)*180/np.pi -np.arctan(tan_value)*180/np.pi)
                duration = time.time() -start_batch
                if angle_diff < 30. and iou >= 0.25:
                    count+=1
                    print('image: %d | duration = %.2f | count = %d | iou = %.2f | angle_difference = %.2f' %(step, duration, count, iou, angle_diff))
            step +=1
    except tf.errors.OutOfRangeError:
        print('Done training for %d epochs, %d steps, %.1f min.' % (FLAGS.num_epochs, step, (time.time()-start_time)/60))
    finally:
        coord.request_stop()

    coord.join(threads)
    sess.close()

def main(_):
    run_training()
    
if __name__ == '__main__':
    parser = argparse.ArgumentParser()
    parser.add_argument(
        '--learning_rate',
        type=float,
        default=0.001,
        help='Initial learning rate.'
    )
    parser.add_argument(
        '--data_dir',
        type=str,
        default='/root/imagenet-data',
        help='Directory with training data.'
    )
    parser.add_argument(
        '--num_epochs',
        type=int,
        default=None,
        help='Number of epochs to run trainer.'
    )
    parser.add_argument(
        '--batch_size',
        type=int,
        default=64,
        help='Batch size.'
    )
    parser.add_argument(
        '--log_dir',
        type=str,
        default='/tmp/tf',
        help='Tensorboard log_dir.'
    )
    parser.add_argument(
        '--model_path',
        type=str,
        default='/tmp/tf/model.ckpt',
        help='Variables for the model.'
    )
    parser.add_argument(
        '--train_or_validation',
        type=str,
        default='validation',
        help='Train or evaluate the dataset'
    )
    FLAGS, unparsed = parser.parse_known_args()
    tf.app.run(main=main, argv=[sys.argv[0]] + unparsed)