Shahab Sabahi's repositories
my-voice-analysis
My-Voice Analysis is a Python library for the analysis of voice (simultaneous speech, high entropy) without the need of a transcription. It breaks utterances and detects syllable boundaries, fundamental frequency contours, and formants.
Machine-Learning-Algorithm-for-Voice-Analysis
It is an algorithm analysed the acoustic features of a voice and creates an acoustic classifier - USEFUL for auto-speech-rater
Neural-Networks-for-Voice-Analysis
It is an algorithm analysed the acoustic features of a voice and creates an acoustic classifier - USEFUL for auto-speech-rater
Converter-pdf-files-to-.txt-or-.html
PDFs are notoriously difficult to scrape. This program converts them to *.txt or *.html formats. The program has tested for Latin alphabets and Japanese.
Legal-Document-Summarization
A comparative study of summarization algorithms applied to legal case judgments by Paheli Bhattacharya
myspokenlanguageid
Spoken language identification with CNN and RNN - Early Stage
mystracher
A structure / Grammar checker of English - Built on AtD, After the Deadline for Python 3.7
Convert-Compare-Images-to-DWG-format
Converting a raster or a vector image to DWG
SUN_TORY_ML_Shelf_Allocation_Project
PCA, Correlation, + Linear/Non-Linear Algorithms: Logistic Regression, Linear Discriminant Analysis, Gaussian Naive Bayes. Support Vector Machines, K-Nearest Neighbors,. Classification and Regression Trees.
Repo-KCIN3D-reconstruction
For the prototype KMCIN1 and feasibility study Feb-March 2021
a-Mini-ASR-Speaker-Identifier
A prototype + an experimental Speech-to-Text Model (modified ver. of A. Pal)
Auto-GPT
An experimental open-source attempt to make GPT-4 fully autonomous.
Complex_Documents_DX
For ideation and experiments
Emotion-classification-from-a-multimodal-Speech-Body-Joint-Movments-dataset
A preliminary experiment setup for Independent-language mental status - classification
Exp_Fraud_Complaint_Analysis
Set up an experiment to compare supervised to unsupervised (done by my colleague) Fraud-Complaint text analysis deploying BERT model. No doubt that the supervised learning was improved using the labeled data, the supervised model has better Precision and Recall (0.86, 0.91) on true positives (fraud cases) than the Unsupervised one (0.52 and 0.72 respectively).
FewShotNLP
The source codes of the paper "Improving Few-shot Text Classification via Pretrained Language Representations" and "When Low Resource NLP Meets Unsupervised Language Model: Meta-pretraining Then Meta-learning for Few-shot Text Classification".
M5-methods
Data, Benchmarks, and methods submitted to the M5 forecasting competition
meta-learning-bert
Meta learning with BERT as a learner
my-Assistant
It measures readability of a written document, and helps to shorten a long written document and to extract key words
myspskills
to predict spoken languages proficiency skills - a Machine Learning algorithm
PPE-Detection-YOLO-Deep_SORT
Real-time PPE detection and tracking using YOLO v3 and deep_sort
Rater-OSAKA-Reading
Reading-Rater for practicing prosody
Sab-AI-Lab
Sab-AI Lab is a boutique AI and machine learning lab in Nagoya-Japan.
Speech-Prosody-Analysis-App
This app consists of a series of Matlab routines useful for various aspects of prosodic analysis of speech.
Spoken-Language-Proiciency-Asssessment-BASIC-VER0
The first step to create an Automatic Spoken Assessment system (Prosody focus)