[FFmpeg-cvslog] dnn_backend_native_layer_mathunary: add atanh support

Ting Fu git at videolan.org
Mon Jul 6 09:45:08 EEST 2020


ffmpeg | branch: master | Ting Fu <ting.fu at intel.com> | Mon Jun 29 22:54:10 2020 +0800| [c0cdeea0ee2c4af9fb4948fe3e33b857cf6d2771] | committer: Guo, Yejun

dnn_backend_native_layer_mathunary: add atanh support

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))

Signed-off-by: Ting Fu <ting.fu at intel.com>

> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=c0cdeea0ee2c4af9fb4948fe3e33b857cf6d2771
---

 libavfilter/dnn/dnn_backend_native_layer_mathunary.c | 4 ++++
 libavfilter/dnn/dnn_backend_native_layer_mathunary.h | 1 +
 tools/python/convert_from_tensorflow.py              | 2 +-
 tools/python/convert_header.py                       | 2 +-
 4 files changed, 7 insertions(+), 2 deletions(-)

diff --git a/libavfilter/dnn/dnn_backend_native_layer_mathunary.c b/libavfilter/dnn/dnn_backend_native_layer_mathunary.c
index b77b84a794..c83d50db64 100644
--- a/libavfilter/dnn/dnn_backend_native_layer_mathunary.c
+++ b/libavfilter/dnn/dnn_backend_native_layer_mathunary.c
@@ -124,6 +124,10 @@ int dnn_execute_layer_math_unary(DnnOperand *operands, const int32_t *input_oper
         for (int i = 0; i < dims_count; ++i)
             dst[i] = acosh(src[i]);
         return 0;
+    case DMUO_ATANH:
+        for (int i = 0; i < dims_count; ++i)
+            dst[i] = atanh(src[i]);
+        return 0;
     default:
         return -1;
     }
diff --git a/libavfilter/dnn/dnn_backend_native_layer_mathunary.h b/libavfilter/dnn/dnn_backend_native_layer_mathunary.h
index eb30231549..8076356ba4 100644
--- a/libavfilter/dnn/dnn_backend_native_layer_mathunary.h
+++ b/libavfilter/dnn/dnn_backend_native_layer_mathunary.h
@@ -42,6 +42,7 @@ typedef enum {
     DMUO_TANH = 9,
     DMUO_ASINH = 10,
     DMUO_ACOSH = 11,
+    DMUO_ATANH = 12,
     DMUO_COUNT
 } DNNMathUnaryOperation;
 
diff --git a/tools/python/convert_from_tensorflow.py b/tools/python/convert_from_tensorflow.py
index 1e73e3aefe..85db7bf710 100644
--- a/tools/python/convert_from_tensorflow.py
+++ b/tools/python/convert_from_tensorflow.py
@@ -72,7 +72,7 @@ class TFConverter:
         self.conv2d_scopename_inputname_dict = {}
         self.op2code = {'Conv2D':1, 'DepthToSpace':2, 'MirrorPad':3, 'Maximum':4, 'MathBinary':5, 'MathUnary':6}
         self.mathbin2code = {'Sub':0, 'Add':1, 'Mul':2, 'RealDiv':3, 'Minimum':4}
-        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}
+        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}
         self.mirrorpad_mode = {'CONSTANT':0, 'REFLECT':1, 'SYMMETRIC':2}
         self.name_operand_dict = {}
 
diff --git a/tools/python/convert_header.py b/tools/python/convert_header.py
index 8fc3438552..9851d84144 100644
--- a/tools/python/convert_header.py
+++ b/tools/python/convert_header.py
@@ -23,4 +23,4 @@ str = 'FFMPEGDNNNATIVE'
 major = 1
 
 # increase minor when we don't have to re-convert the model file
-minor = 17
+minor = 18



More information about the ffmpeg-cvslog mailing list