Update README and misc changes

JDAI-CV · Jun 5, 2019 · 26b4fcf · 26b4fcf
1 parent 0f0c183
commit 26b4fcf
Show file tree

Hide file tree

Showing 3 changed files with 7 additions and 3 deletions.
diff --git a/README.md b/README.md
@@ -62,7 +62,9 @@ We publish two pretrained binary neural network models based on [Bi-Real Net](ht
 
 ## Implementation Details
 
-We plan to participate the [ACM Multimedia 2019 Open Source Software Competition](https://www.acmmm.org/2019/call-for-open-source-software-competition/). Our implementation details will be presented in a 4-page short paper soon.
+* The Implementation of Binary Convolutions: [docs/bconv.md](docs/bconv.md)
+
+* Model Conversion: [docs/onnx2bnn.md](docs/onnx2bnn.md)
 
 ## Example project
 

diff --git a/README_CN.md b/README_CN.md
@@ -64,7 +64,9 @@ dabnn_bireal18_imagenet_stem   43279353 ns   41533009 ns         14       <---
 
 ## 技术细节
 
-我们计划参加 [ACM Multimedia 2019 Open Source Software Competition](https://www.acmmm.org/2019/call-for-open-source-software-competition/). dabnn 的技术细节很快会在一篇四页的短论文中描述。
+* Binary Convolutions 的实现: [docs/bconv.md](docs/bconv_CN.md)
+
+* 模型转换: [docs/onnx2bnn.md](docs/onnx2bnn_CN.md)
 
 ## 示例工程
 

diff --git a/docs/overall.md b/docs/overall.md
@@ -2,7 +2,7 @@
 
 Binary Neural Networks is proposed in [Binary Neural Networks](https://arxiv.org/abs/1602.02830) and [XNOR-Net](https://arxiv.org/abs/1603.05279). In the following papers, [Bi-Real Net](https://arxiv.org/abs/1808.00278) presented some new training method in order to improve the performance, [BENN](https://arxiv.org/abs/1806.07550) leverages emsemble on BNNs.
 
-BNNs can save 10X+ memory, and several times as fast as float NNs. What's more, it theoretically [saves 10X energy](https://camo.githubusercontent.com/e725038be60ce4bb698b22480603b636a92beeaf/687474703a2f2f66696c652e656c656366616e732e636f6d2f776562312f4d30302f35352f37392f7049594241467373565f5341504f63534141435742546f6d6531633033392e706e67). So the battery life of devices will be expanded a lot.
+BNNs can save 10X+ memory, and several times as fast as float NNs. What's more, it theoretically [saves 10X energy](https://camo.githubusercontent.com/e725038be60ce4bb698b22480603b636a92beeaf/687474703a2f2f66696c652e656c656366616e732e636f6d2f776562312f4d30302f35352f37392f7049594241467373565f5341504f63534141435742546f6d6531633033392e706e67), so the battery life of devices will be expanded a lot.
 
 ## Some notes