This page is the collections of my writting samples, including project reports, presentations, and paper reviews. Most of the papers are written from 2019-2020.
I am really skilled at designing and modeling with Solidworks. I am very familiar with using UltimakerCura and using 3D printing.
I can use NI Multisim and EasyEda for PCB design and layout. I am very skilled with testing equipment like multimeter, function generator, and oscilloscope. Strong hands-on skills with soldering.
The prototyping flow for this speaker circuitry:
Design, testing, and simulation with NI Multisim->
PCB design, PCB layout, PCB production (outsource) ->
Soldering components and do testing!
This document is how I use EasyEDA to create schemetics and PCBs.
This section contains the reports on multiple projects I did in the past few years, for detailed project information, please visit the Project section.
Music Generation with GANs
Multi_Speaker_TTS_Reviews_and_Implementation
Bio-gaming for abelow-elbow amputee
Website filter for WLAN
Recurrent_neural_network_based_language_model_reading
Feature Extraction MFCC
END_TO_END_ATTENTION_BASED_LARGE_VOCABULARY_SPEECH_RECOGNITION
Deep NeuralNets for Acoustic modeling
Acoustic modeling based on the mdl principle for speech recognition.pdf
Conversational_Speech_Transcription_Using_Context_Dependent_Deep_Neural_Networks_information_extraction
Applying convolutional neural networks concepts to hybrid NN-HMM model for Speech recognition
A Gentle Tutorial of the EM Algorithm and its Application to Parameter Estimation for Gaussian Mixture and Hidden Markov Models
Tree_based state tying for high accuracy acoustic modeling
WAVENET—— A GENERATIVE MODEL FOR RAW AUDIO
K-means clustering & Gaussian Mixture model presentation
Comparison of END-to-END, CTC, Attention based Encoder-Decoder system for speech recognition
AlexNet
HOGgles_visualizing object detection features
Resnet
Robust Face Recognition via Sparse Representation
Eigenfaces and fisherfaces
Face Recognition Using Eigen Faces
Dropout, a Simple way to prevent NN from overfitting
Batch_Normalization__Accelerating_Deep_Network_Training_b_y_Reducing_Internal_Covariate_Shift
KSVD_An_Algorithm_for_Designing_Overcomplete_Dictionaries_for_Sparse_Representation
going_deeper_with_convolutions
I wrote this technical documentation for our controllers.
I made this interactive conditional formated controller picking website: https://ecotron.ai/pick-your-controllers/