It can be tested with the model generated with below python script: import tensorflow as tf import numpy as np import imageio in_img = imageio.imread('input.jpeg') in_img = in_img.astype(np.float32)/255.0 in_data = in_img[np.newaxis, :] x = tf.placeholder(tf.float32, shape=[1, None, None, 3], name='dnn_in') please uncomment the part you want to test x_sinh_1 = tf.sinh(x) x_out = tf.divide(x_sinh_1, 1.176) # sinh(1.0) x_cosh_1 = tf.cosh(x) x_out = tf.divide(x_cosh_1, 1.55) # cosh(1.0) x_tanh_1 = tf.tanh(x) x__out = tf.divide(x_tanh_1, 0.77) # tanh(1.0) x_asinh_1 = tf.asinh(x) x_out = tf.divide(x_asinh_1, 0.89) # asinh(1.0/1.1) x_acosh_1 = tf.add(x, 1.1) x_acosh_2 = tf.acosh(x_acosh_1) # accept (1, inf) x_out = tf.divide(x_acosh_2, 1.4) # acosh(2.1) x_atanh_1 = tf.divide(x, 1.1) x_atanh_2 = tf.atanh(x_atanh_1) # accept (-1, 1) x_out = tf.divide(x_atanh_2, 1.55) # atanhh(1.0/1.1) y = tf.identity(x_out, name='dnn_out') #please only preserve the x_out you want to test sess=tf.Session() sess.run(tf.global_variables_initializer()) graph_def = tf.graph_util.convert_variables_to_constants(sess, sess.graph_def, ['dnn_out']) tf.train.write_graph(graph_def, '.', 'image_process.pb', as_text=False) print("image_process.pb generated, please use \ path_to_ffmpeg/tools/python/convert.py to generate image_process.model\n") output = sess.run(y, feed_dict={x: in_data}) imageio.imsave("out.jpg", np.squeeze(output))
Unified · Split
| @@ -124,6 +124,10 @@ int dnn_execute_layer_math_unary(DnnOperand *operands, const int32_t *input_oper · 123 unmodified lines | |||
| 124 | for (int i = 0; i < dims_count; ++i) | 124 | for (int i = 0; i < dims_count; ++i) |
| 125 | dst[i] = acosh(src[i]); | 125 | dst[i] = acosh(src[i]); |
| 126 | return 0; | 126 | return 0; |
| 127 | case DMUO_ATANH: | ||
| 128 | for (int i = 0; i < dims_count; ++i) | ||
| 129 | dst[i] = atanh(src[i]); | ||
| 130 | return 0; | ||
| 127 | default: | 131 | default: |
| 128 | return -1; | 132 | return -1; |
| 129 | } | 133 | } |
| @@ -42,6 +42,7 @@ typedef enum { · 41 unmodified lines | |||
| 42 | DMUO_TANH = 9, | 42 | DMUO_TANH = 9, |
| 43 | DMUO_ASINH = 10, | 43 | DMUO_ASINH = 10, |
| 44 | DMUO_ACOSH = 11, | 44 | DMUO_ACOSH = 11, |
| 45 | DMUO_ATANH = 12, | ||
| 45 | DMUO_COUNT | 46 | DMUO_COUNT |
| 46 | } DNNMathUnaryOperation; | 47 | } DNNMathUnaryOperation; |
| 47 | 48 | ||
| @@ -72,7 +72,7 @@ class TFConverter: · 71 unmodified lines | |||
| 72 | self.conv2d_scopename_inputname_dict = {} | 72 | self.conv2d_scopename_inputname_dict = {} |
| 73 | self.op2code = {'Conv2D':1, 'DepthToSpace':2, 'MirrorPad':3, 'Maximum':4, 'MathBinary':5, 'MathUnary':6} | 73 | self.op2code = {'Conv2D':1, 'DepthToSpace':2, 'MirrorPad':3, 'Maximum':4, 'MathBinary':5, 'MathUnary':6} |
| 74 | self.mathbin2code = {'Sub':0, 'Add':1, 'Mul':2, 'RealDiv':3, 'Minimum':4} | 74 | self.mathbin2code = {'Sub':0, 'Add':1, 'Mul':2, 'RealDiv':3, 'Minimum':4} |
| 75 | self.mathun2code = {'Abs':0, 'Sin':1, 'Cos':2, 'Tan':3, 'Asin':4, 'Acos':5, 'Atan':6, 'Sinh':7, 'Cosh':8, 'Tanh':9, 'Asinh':10, 'Acosh':11} | 75 | self.mathun2code = {'Abs':0, 'Sin':1, 'Cos':2, 'Tan':3, 'Asin':4, 'Acos':5, 'Atan':6, 'Sinh':7, 'Cosh':8, 'Tanh':9, 'Asinh':10, 'Acosh':11, 'Atanh':12} |
| 76 | self.mirrorpad_mode = {'CONSTANT':0, 'REFLECT':1, 'SYMMETRIC':2} | 76 | self.mirrorpad_mode = {'CONSTANT':0, 'REFLECT':1, 'SYMMETRIC':2} |
| 77 | self.name_operand_dict = {} | 77 | self.name_operand_dict = {} |
| 78 | 78 | ||
| @@ -23,4 +23,4 @@ str = 'FFMPEGDNNNATIVE' · 22 unmodified lines | |||
| 23 | major = 1 | 23 | major = 1 |
| 24 | 24 | ||
| 25 | # increase minor when we don't have to re-convert the model file | 25 | # increase minor when we don't have to re-convert the model file |
| 26 | minor = 17 | 26 | minor = 18 |