diff --git a/Documentation/current_iteration.md b/Documentation/current_iteration.md index dab501987a49..0a476c8f7834 100644 --- a/Documentation/current_iteration.md +++ b/Documentation/current_iteration.md @@ -199,6 +199,9 @@ We added support for dilation convolution on the GPU, exposed by BrainScript, C+ ### Deterministic Pooling Now call `cntk.debug.force_deterministic()` will make max and average pooling deterministic, this behavior depend on cuDNN version 6 or later. +### Add Crop Node to Python API +Motivation in order to support some image segmentation network, we added Crop node to C++ and Python API. Crop node crops its first input along spatial axes so that the result matches the spatial size of its second (reference) input. All non-spatial dimensions are unchanged. Crop offsets can be specified directly, or computed automatically, by traversing the network, and matching centers of receptive fields between activations in the two inputs + ## Performance ### Intel MKL update to improve inference speed on CPU by around 2x on AlexNet This work is rolled over to next release due to dependency on test infrastructure updates. @@ -215,16 +218,25 @@ We added the image feature support for TensorBoard. Now CNTK users can use Tenso ### Acknowledgments We thank the following community members for their contributions:+ +* [@arturl](https://github.com/arturl) +* [@chentaMS](https://github.com/chentaMS) * [@cijianzy](https://github.com/cijianzy) +* [@DaveyBiggers](https://github.com/DaveyBiggers) * [@duli2012](https://github.com/duli2012) +* [@eldakms](https://github.com/eldakms) * [@FDecaYed](https://github.com/FDecaYed) +* [@junjieqian](https://github.com/junjieqian) * [@karolzak](https://github.com/karolzak) * [@KGWANGMIN](https://github.com/KGWANGMIN) +* [@mnidza](https://github.com/mnidza) * [@ottolu](https://github.com/ottolu) +* [@raaaar](https://github.com/raaaar) * [@StillKeepTry](https://github.com/StillKeepTry) * [@taehoonlee](https://github.com/taehoonlee) +* [@vmazalov](https://github.com/vmazalov) + We apologize for any community contributions we might have overlooked in these release notes. ## Others -### Continue work on [Deep Learning Explained](https://www.edx.org/course/deep-learning-explained-microsoft-dat236x) course on edX. +### Continue work on [Deep Learning Explained](https://www.edx.org/course/deep-learning-explained-microsoft-dat236x) course on edX. \ No newline at end of file