Skip to content

Commit 0a54510

Browse files
authored
Update FEATURES.md
1 parent 9aec595 commit 0a54510

File tree

1 file changed

+19
-9
lines changed

1 file changed

+19
-9
lines changed

FEATURES.md

Lines changed: 19 additions & 9 deletions
Original file line numberDiff line numberDiff line change
@@ -3,12 +3,12 @@ A Short Summary of New Features in Synopsys Caffe
33

44
Synopsys Caffe Version: 2018.09
55
New added features are compared with the original BVLC Caffe 1.0.0
6-
6+
77
evconvert (TensorFlow/ONNX/... to Caffe Converter) related
8-
------------------------------
9-
pad_type ("SAME" style padding) in ConvolutionParameter and PoolingParameter
8+
----------------------------------------------------------
9+
pad_type (deprecated, "SAME" style padding) in ConvolutionParameter and PoolingParameter
1010
pad_l, pad_r, pad_t and pad_b (arbitrary 2D padding) in ConvolutionParameter and PoolingParameter
11-
AVE_EXC_PAD (average pooling excluding the paddings) in PoolingParameter
11+
AVE_EXC_PAD (average pooling excluding the paddings), AVE_TF (deprecated, alias for AVE_EXC_PAD) in PoolingParameter
1212
ceil_mode in PoolingParameter
1313
relu6 in ReLUParameter
1414
pad.py (customized Python layer)
@@ -21,6 +21,20 @@ squeeze_conv_layer
2121
squeeze_inner_product_layer
2222
squeeze_deconv_layer
2323

24+
YOLOv2 related
25+
--------------
26+
add_eps_before_sqrt in BatchNormParameter, MVNParameter and NormalizeParameter
27+
caffe_yolo in TransformationParameter
28+
jitter in ResizeParameter
29+
exposure_lower, exposure_upper in DistortionParameter
30+
side in DataParameter
31+
reorg_layer
32+
yolo_v2_loss_layer
33+
34+
model reference acceleration related
35+
------------------------------------
36+
input_scale, output_scale in ConvolutionParameter and InnerProductParameter
37+
2438
ICNet (PSPNet) related
2539
---------------------
2640
adaptive_bias_channel_layer
@@ -39,11 +53,7 @@ unique_label_layer
3953
bn_layer (slope_filler, bias_filler, momentum and icnet in BNParameter)
4054
update_global_stats and icnet in BatchNormParameter
4155
scale_factors, crop_width and crop_height in TransformationParameter
42-
43-
YOLOv2 related
44-
--------------
45-
add_eps_before_sqrt in BatchNormParameter, MVNParameter and NormalizeParameter
46-
56+
4757
SRGAN related
4858
-------------
4959
gan_loss_layer

0 commit comments

Comments
 (0)