1D convolutions are extensively used in digital signal processing (filtering/denoising) and analysis (also through CNNs). As their computational complexity is of the order O(N^2), their fast execution is a must. This lecture will overview linear systems, linear and cyclic convolution and correlation. Then it will present their fast execution through FFTs, resulting in algorithms having computational… Continue reading Fast 1D Convolution Algorithms
Archives: Resources
Resource description
Digital Filter Structures
This lecture overviews Digital Filter Structures that has many applications in digital signal processing. It covers the following topics in detail: IIR Filter Structures (Direct Filter Structure, Cascade Filter Structure, Parallel Filter Structure, Transposed Filter Structure). FIR Filter Structures (Direct Filter Structure, Cascade Filter Structure, Parallel Filter Structure, Frequency Sampling Structure).
FIR Filter Design
This lecture overviews FIR Filter Design that has many applications in digital signal processing and deep neural networks. It covers the following topics in detail: Window Method, Optimization Methods, Frequency Sampling Method, Equiripple FIR Filter Design.
NLP and Text Sentiment Analysis
This lecture overviews Natural Language Processing (NLP) and Text Sentiment Analysis that has many applications in Text Analytics, Opinion extraction, Opinion mining, Sentiment mining, Subjectivity analysis. It covers the following topics in detail: Baseline algorithms. Text pre-processing. Word embeddings (Word2Vec, Fast Text). Neural NLP and Sentiment Analysis for text classification (RNN, CNN). Contextual Embeddings (ELMo).… Continue reading NLP and Text Sentiment Analysis
Natural Language Processing
This lecture overviews Natural Language Processing (NLP) that has many applications in text analytics, Linguistics, Machine translation and sentiment analysis. It covers the following topics in detail: Symbolic NLP, Statistical NLP, Neural NLP. NLP methods: Rules, Statistics, Neural networks. Word Representations: Fixed (sparse), One-hot encoding, Bag-of-words, TF-IDF Distributed (dense). Classic embeddings: Word2Vec, GloVe, FastText. Contextualized embeddings: CoVe,… Continue reading Natural Language Processing
Neural Speech Recognition
This lecture overviews Neural Speech Recognition is a special case of Automatic Speech Recognition (ASR), i.e., the transcription of speech to text that has many applications e.g., in call centers, dictation, meeting minutes creation, Smart assistants (Apple’s Siri, Amazon’s Alexa, Google Assistant, Microsoft’s Cortana) and in Behavior /emotion recognition. It covers the following topics in… Continue reading Neural Speech Recognition