Reverse input channels fusion (#4276)

* Side fix found while working on Windows machine.

* Fix for non-fused Reverse Input Channels subgraph

model-optimizer/extensions/back/ReverseInputChannels.py

@@ -97,7 +97,9 @@ class ReverseChannelsPropagationDown(BackReplacementPattern):
         'BatchNormalization': lambda node, rc: ReverseChannelsPropagationDown.pass_rc_through_eltwise(node, rc),
         'FakeQuantize': lambda node, rc: ReverseChannelsPropagationDown.pass_rc_through_eltwise(node, rc),
         'Multiply': lambda node, rc: ReverseChannelsPropagationDown.pass_rc_through_eltwise(node, rc),
+        'Divide': lambda node, rc: ReverseChannelsPropagationDown.pass_rc_through_eltwise(node, rc),
         'Add': lambda node, rc: ReverseChannelsPropagationDown.pass_rc_through_eltwise(node, rc),
+        'Subtract': lambda node, rc: ReverseChannelsPropagationDown.pass_rc_through_eltwise(node, rc),
         'Pow': lambda node, rc: ReverseChannelsPropagationDown.pass_rc_through_eltwise(node, rc),
         'Convert': lambda node, rc: ReverseChannelsPropagationDown.pass_rc_through_eltwise(node, rc),
 
@@ -198,9 +200,8 @@ def pass_rc_through_eltwise(node, reverse_channels):
                 continue
             shape = port.data.get_shape()
             non_one_dims = np.where(shape != 1)[0]
-            if len(non_one_dims) == 0:
-                # shape contains only ones - nothing to flip for this input
-                continue
+            if shape[reverse_channels.axis] == 1:
+                continue  # nothing to flip for this input
             if len(non_one_dims) == 1 and shape[non_one_dims.item()] == reverse_channels.order.size:
                 new_axis = non_one_dims.item()
             elif np.array_equal(before_shape, shape):
@@ -238,7 +239,8 @@ def pass_rc_through_shape(node, reverse_channels):
         """
         stops propagation of RIC through shape taking operations, due to RIC does not change shape
         """
-        reverse_channels.out_port(0).get_connection().set_source(reverse_channels.in_port(0).get_connection().get_source())
+        reverse_channels.out_port(0).get_connection().set_source(
+            reverse_channels.in_port(0).get_connection().get_source())
         return False
 
     @staticmethod
@@ -269,7 +271,9 @@ class ReverseChannelsPropagationUp(BackReplacementPattern):
         'BatchNormalization': lambda node, rc: ReverseChannelsPropagationUp.lift_up_through_eltwise(node, rc),
         'FakeQuantize': lambda node, rc: ReverseChannelsPropagationUp.lift_up_through_eltwise(node, rc),
         'Multiply': lambda node, rc: ReverseChannelsPropagationUp.lift_up_through_eltwise(node, rc),
+        'Divide': lambda node, rc: ReverseChannelsPropagationUp.lift_up_through_eltwise(node, rc),
         'Add': lambda node, rc: ReverseChannelsPropagationUp.lift_up_through_eltwise(node, rc),
+        'Subtract': lambda node, rc: ReverseChannelsPropagationUp.lift_up_through_eltwise(node, rc),
         'Pow': lambda node, rc: ReverseChannelsPropagationUp.lift_up_through_eltwise(node, rc),
         'Convert': lambda node, rc: ReverseChannelsPropagationUp.lift_up_through_eltwise(node, rc),
     }
@@ -300,9 +304,8 @@ def lift_up_through_eltwise(node: Node, reverse_channels: Node):
             shape = port.data.get_shape()
 
             non_one_dims = np.where(shape != 1)[0]
-            if len(non_one_dims) == 0:
-                # shape contains only ones - nothing to flip for this input
-                continue
+            if shape[reverse_channels.axis] == 1:
+                continue  # nothing to flip for this input
             if len(non_one_dims) == 1 and shape[non_one_dims.item()] == reverse_channels.order.size:
                 axis = non_one_dims.item()
             elif np.array_equal(before_shape, shape):

model-optimizer/extensions/front/DropoutWithRandomUniformReplacer.py

@@ -20,6 +20,7 @@
 from mo.front.common.replacement import FrontReplacementSubgraph
 from mo.front.tf.graph_utils import create_op_with_const_inputs
 from mo.graph.graph import Graph, Node, rename_nodes
+from mo.middle.pattern_match import check_value
 from mo.ops.broadcast import Broadcast
 
 
@@ -51,7 +52,7 @@ def pattern(**kwargs):
                 ('shape', dict(op='ShapeOf')),
                 ('random_uniform', dict(op='RandomUniform')),
                 ('mul', dict(op='Mul')),
-                ('add_const', dict(op='Const', value=lambda v: v is not None and np.allclose(v, 0.0, atol=0))),
+                ('add_const', dict(op='Const', value=lambda v: check_value(v, lambda x: np.allclose(x, 0.0, atol=0)))),
                 ('add', dict(op='Add')),
                 ('add2', dict(op='Add')),
                 ('floor', dict(op='Floor')),

model-optimizer/extensions/front/HSigmoid_fusion.py

@@ -20,6 +20,7 @@
 from mo.front.common.replacement import FrontReplacementSubgraph
 from mo.front.subgraph_matcher import SubgraphMatch
 from mo.graph.graph import Graph, rename_nodes
+from mo.middle.pattern_match import check_value
 from mo.utils.graph import Node
 
 
@@ -50,10 +51,11 @@ def pattern(self):
             nodes=[
                 ('input', dict()),
                 ('add', dict(op='Add')),
-                ('const_0', dict(op='Const', value=lambda v: v is not None and np.allclose(v, 0.0, atol=1e-6))),
-                ('const_3', dict(op='Const', value=lambda v: v is not None and np.allclose(v, 3.0, atol=1e-6))),
-                ('const_6', dict(op='Const', value=lambda v: v is not None and np.allclose(v, 6.0, atol=1e-6))),
-                ('const_1_6', dict(op='Const', value=lambda v: v is not None and np.allclose(v, 1.0 / 6.0, atol=1e-6))),
+                ('const_0', dict(op='Const', value=lambda v: check_value(v, lambda x: np.allclose(x, 0.0, atol=1e-6)))),
+                ('const_3', dict(op='Const', value=lambda v: check_value(v, lambda x: np.allclose(x, 3.0, atol=1e-6)))),
+                ('const_6', dict(op='Const', value=lambda v: check_value(v, lambda x: np.allclose(x, 6.0, atol=1e-6)))),
+                ('const_1_6',
+                 dict(op='Const', value=lambda v: check_value(v, lambda x: np.allclose(x, 1.0 / 6.0, atol=1e-6)))),
                 ('clamp', dict(op='Clamp')),
                 ('mul_2', dict(op='Mul')),
             ],
@@ -86,10 +88,11 @@ def pattern(self):
             nodes=[
                 ('input', dict()),
                 ('add', dict(op='Add')),
-                ('const_0', dict(op='Const', value=lambda v: v is not None and np.allclose(v, 0.0, atol=1e-6))),
-                ('const_3', dict(op='Const', value=lambda v: v is not None and np.allclose(v, 3.0, atol=1e-6))),
-                ('const_6', dict(op='Const', value=lambda v: v is not None and np.allclose(v, 6.0, atol=1e-6))),
-                ('const_1_6', dict(op='Const', value=lambda v: v is not None and np.allclose(v, 1.0 / 6.0, atol=1e-6))),
+                ('const_0', dict(op='Const', value=lambda v: check_value(v, lambda x: np.allclose(x, 0.0, atol=1e-6)))),
+                ('const_3', dict(op='Const', value=lambda v: check_value(v, lambda x: np.allclose(x, 3.0, atol=1e-6)))),
+                ('const_6', dict(op='Const', value=lambda v: check_value(v, lambda x: np.allclose(x, 6.0, atol=1e-6)))),
+                ('const_1_6',
+                 dict(op='Const', value=lambda v: check_value(v, lambda x: np.allclose(x, 1.0 / 6.0, atol=1e-6)))),
                 ('max', dict(op='Maximum')),
                 ('min', dict(op='Minimum')),
                 ('mul_2', dict(op='Mul')),
@@ -123,12 +126,15 @@ def pattern(self):
         return dict(
             nodes=[
                 ('input', dict()),
-                ('add_const', dict(op='Const', value=lambda v: v is not None and np.allclose(v, 3.0, atol=1e-6))),
+                ('add_const',
+                 dict(op='Const', value=lambda v: check_value(v, lambda x: np.allclose(x, 3.0, atol=1e-6)))),
                 ('add', dict(op='Add')),
                 ('relu', dict(op='ReLU')),
-                ('min_const', dict(op='Const', value=lambda v: v is not None and np.allclose(v, 6.0, atol=1e-6))),
+                ('min_const',
+                 dict(op='Const', value=lambda v: check_value(v, lambda x: np.allclose(x, 6.0, atol=1e-6)))),
                 ('min', dict(op='Minimum')),
-                ('div_const', dict(op='Const', value=lambda v: v is not None and np.allclose(v, 6.0, atol=1e-6))),
+                ('div_const',
+                 dict(op='Const', value=lambda v: check_value(v, lambda x: np.allclose(x, 6.0, atol=1e-6)))),
                 ('div', dict(op='Div')),
             ],
             edges=[
@@ -159,12 +165,15 @@ def pattern(self):
         return dict(
             nodes=[
                 ('input', dict()),
-                ('add_const', dict(op='Const', value=lambda v: v is not None and np.allclose(v, 3.0, atol=1e-6))),
+                ('add_const',
+                 dict(op='Const', value=lambda v: check_value(v, lambda x: np.allclose(x, 3.0, atol=1e-6)))),
                 ('add', dict(op='Add')),
                 ('relu', dict(op='ReLU')),
-                ('min_const', dict(op='Const', value=lambda v: v is not None and np.allclose(v, 6.0, atol=1e-6))),
+                ('min_const',
+                 dict(op='Const', value=lambda v: check_value(v, lambda x: np.allclose(x, 6.0, atol=1e-6)))),
                 ('min', dict(op='Minimum')),
-                ('mul_const', dict(op='Const', value=lambda v: v is not None and np.allclose(v, 1.0/6.0, atol=1e-6))),
+                ('mul_const',
+                 dict(op='Const', value=lambda v: check_value(v, lambda x: np.allclose(x, 1.0 / 6.0, atol=1e-6)))),
                 ('mul', dict(op='Mul')),
             ],
             edges=[

model-optimizer/extensions/front/HSwish_fusion.py

@@ -20,6 +20,7 @@
 from mo.front.common.replacement import FrontReplacementSubgraph
 from mo.front.subgraph_matcher import SubgraphMatch
 from mo.graph.graph import Graph, rename_nodes
+from mo.middle.pattern_match import check_value
 
 
 def replace_with_hswish(graph: Graph, match: [dict, SubgraphMatch]):
@@ -58,10 +59,11 @@ def pattern(self):
             nodes=[
                 ('input', dict()),
                 ('add', dict(op='Add')),
-                ('const_0', dict(op='Const', value=lambda v: v is not None and np.allclose(v, 0.0, atol=1e-6))),
-                ('const_3', dict(op='Const', value=lambda v: v is not None and np.allclose(v, 3.0, atol=1e-6))),
-                ('const_6', dict(op='Const', value=lambda v: v is not None and np.allclose(v, 6.0, atol=1e-6))),
-                ('const_1_6', dict(op='Const', value=lambda v: v is not None and np.allclose(v, 1 / 6.0, atol=1e-6))),
+                ('const_0', dict(op='Const', value=lambda v: check_value(v, lambda x: np.allclose(x, 0.0, atol=1e-6)))),
+                ('const_3', dict(op='Const', value=lambda v: check_value(v, lambda x: np.allclose(x, 3.0, atol=1e-6)))),
+                ('const_6', dict(op='Const', value=lambda v: check_value(v, lambda x: np.allclose(x, 6.0, atol=1e-6)))),
+                ('const_1_6',
+                 dict(op='Const', value=lambda v: check_value(v, lambda x: np.allclose(x, 1.0 / 6.0, atol=1e-6)))),
                 ('clamp', dict(op='Clamp')),
                 ('mul', dict(op='Mul')),
                 ('mul_2', dict(op='Mul')),
@@ -97,10 +99,11 @@ def pattern(self):
             nodes=[
                 ('input', dict()),
                 ('add', dict(op='Add')),
-                ('const_0', dict(op='Const', value=lambda v: v is not None and np.allclose(v, 0.0, atol=1e-6))),
-                ('const_3', dict(op='Const', value=lambda v: v is not None and np.allclose(v, 3.0, atol=1e-6))),
-                ('const_6', dict(op='Const', value=lambda v: v is not None and np.allclose(v, 6.0, atol=1e-6))),
-                ('const_1_6', dict(op='Const', value=lambda v: v is not None and np.allclose(v, 1 / 6.0, atol=1e-6))),
+                ('const_0', dict(op='Const', value=lambda v: check_value(v, lambda x: np.allclose(x, 0.0, atol=1e-6)))),
+                ('const_3', dict(op='Const', value=lambda v: check_value(v, lambda x: np.allclose(x, 3.0, atol=1e-6)))),
+                ('const_6', dict(op='Const', value=lambda v: check_value(v, lambda x: np.allclose(x, 6.0, atol=1e-6)))),
+                ('const_1_6',
+                 dict(op='Const', value=lambda v: check_value(v, lambda x: np.allclose(x, 1.0 / 6.0, atol=1e-6)))),
                 ('max', dict(op='Maximum')),
                 ('min', dict(op='Minimum')),
                 ('mul', dict(op='Mul')),

model-optimizer/extensions/front/Softplus_fusion.py

@@ -19,6 +19,7 @@
 from mo.front.common.replacement import FrontReplacementSubgraph
 from mo.front.subgraph_matcher import SubgraphMatch
 from mo.graph.graph import Graph, rename_nodes
+from mo.middle.pattern_match import check_value
 
 
 class SoftplusFusion(FrontReplacementSubgraph):
@@ -32,7 +33,7 @@ def pattern(self):
             nodes=[
                 ('exp', dict(op='Exp')),
                 ('add', dict(op='Add')),
-                ('const_1', dict(op='Const', value=lambda v: v is not None and np.allclose(v, 1.0, atol=1e-6))),
+                ('const_1', dict(op='Const', value=lambda v: check_value(v, lambda x: np.allclose(x, 1.0, atol=1e-6)))),
                 ('ln', dict(op='Log')),
             ],
             edges=[

model-optimizer/mo/middle/pattern_match.py

@@ -16,6 +16,7 @@
 
 import logging as log
 
+import numpy as np
 from networkx.algorithms import isomorphism as ism
 
 from mo.graph.graph import Node, dict_includes, Graph
@@ -163,3 +164,7 @@ def find_isomorphisms(graph: Graph, nodes: list, edges: list):
         match = {k: Node(graph, match[k]) for k in match.keys()}
         result.append(match)
     return result
+
+
+def check_value(v: np.ndarray, check: callable):
+    return v is not None and np.all(np.isreal(v)) and check(v)