A TensorFlow implementation of Local Binary Patterns Transform
Feature extraction layers like Local Binary Patterns (LBP) transform can be very useful for improving the accuracy of machine learning and deep learning models depending on the problem type. Direct implementations of such layers in Python may result in long running times, and training a computer vision model may be delayed significantly. For this purpose, TensorFlow framework enables developing accelerated custom operations based on the existing operations which already have support for accelerated hardware such as multicore CPU and GPU. In this study, LBP transform which is used for feature extraction in various applications, was implemented based on TensorFlow operations. The evaluations were done using both standard Python operations and TensorFlow library for performance comparisons. The experiments were realized using images in various dimensions and various batch sizes. Numerical results show that algorithm based on TensorFlow operations provides good acceleration rates over Python runs. The implementation of LBP can be used for the accelerated computing for various feature extraction purposes including machine learning as well as in deep learning applications.
Code (1)
Tasks
CPUGPUSimilar Papers 제목 키워드 기반
Larq Compute Engine: Design, Benchmark, and Deploy State-of-the-Art Binarized Neural Networks
We introduce Larq Compute Engine, the world's fastest Binarized Neural Network (BNN) inference engine, and use this framework to investigate several important questions about the efficiency of BNNs and to design a new st…
TensorFlow based deep learning layer for Local Derivative Patterns
Deep learning architectures can extract features automatically via particular layers such as convolutions, recurrent networks, and pooling operations. In addition to the predefined layers in TensorFlow, developers can im…
Deep LearningLightSeq: A High Performance Inference Library for Transformers
Transformer, BERT and their variants have achieved great success in natural language processing. Since Transformer models are huge in size, serving these models is a challenge for real industrial applications. In this pa…
GPUMachine TranslationTranslationVocal Bursts Intensity PredictionShould I use TensorFlow
Google's Machine Learning framework TensorFlow was open-sourced in November 2015 [1] and has since built a growing community around it. TensorFlow is supposed to be flexible for research purposes while also allowing its …
BIG-bench Machine LearningGeneral ClassificationEfficient Sampled Softmax for Tensorflow
This short paper discusses an efficient implementation of \emph{sampled softmax loss} for Tensorflow. The speedup over the default implementation is achieved due to simplification of the graph for the forward and backwar…