Voice Emotion Detection
Detecting emotion in speech using Neural Networks with trainable Mel Frequency Cepstrum Coefficients (MFFC) layers.
Detecting emotion in speech using Neural Networks with trainable Mel Frequency Cepstrum Coefficients (MFFC) layers.
Recognizing a person's gender in real time using Gaussian Mixture Models.
Training various Machine Learning models to detect both gender and emotion from Mel Frequency Cepstrum Coefficients (MFFC) extracted from audio.
Training a Neural Network to selectively apply extractors to text for automated information mining.
Employing web scraping techniques to filter and assess the speeches of NY Governor Andrew Cuomo and NY Mayor Bill de Blasio during the covid-19 pandemic.
Training a neural network to replicate Bert Russell's writing style.
Handcoded implementation of the Transformer encoder architecture as described in the BERT paper.
Leveraging lexical and syntactical text attributes for information extraction on unstructured web data.
Employing web scraping and a Bidirectional Encoder Representations from Transformers (BERT) neural network to analyse customer sentiment on multiple Apps competing in the work management space.
Combining text feature engineering and Bidirectional Encoder Representations from Transformers (BERT) to automatically classify a TED Talk's topic.
Developing a personal voice assistant leveraging existing real-time voice detection libraries.
Translating English to German using a Sequence-to-Sequence neural network.
Applying NLP feature engineering techniques to tweets to create enhanced vector representations of the data, which are used to train a recurrent neural network for disaster prediction.
Employing transfer learning (Xception and GloVe) to automatically extract image and text features used to train a neural network capable of captioning images.
Identifying matching entities in noisy datasets generated from web scraping, then merging the datasets and creating an unifying ontology.
Cross-source entity resolution with Probabilistic Soft Logic (PSL), employing external attributes for collective entity resolution.
Knowlege Graph about AI Research, built via web scraping, entity resolution, external KG linkage, ontology generation and RDF deployment. Includes PageRank, Topic Modeling, Recommender System and UI.
Scraping theme-specific articles from main news outlets, then evaluating overall sentiment through various NLP techniques.
Automating sentiment analysis of Youtube videos with Amazon Comprehend.