Mohammad Arshad  Mohammad Arshad photo         

contact information

AI Research
  +91dash8750dash301292

links

Professional Associations

Professional Associations:  IBM Academy of Technology

profile


Exposure: Having total of 13+ experience with currently working on Neural Network solution using python and Tensor flow, Keras, Caffe, creating models using python machine learning library sklearn and building NLP solutions, model porting on Mobile device and Nvidia Jetson Tx2, expertise on Bluemix Cloud. Major past experience in various J2EE technologies in design and development activities.

 

AI/Data

Recently developed a NPA solution using Tensorflow and Python as part of TEC-I, and published a paper in AOT

http://aot.raleigh.ibm.com/aot/aotrpts.nsf/d780c871ccfebc2b8525768d004e454a/f609b2e9ef2ed3d0852582c5005c4e9d?OpenDocument

https://github.com/arshad2101/CognitiveBanking/blob/master/Cognitive_Banking_MLNN.v2.ipynb

 

Created a CNN model to classify the Diabetic Retinopathy, this a custom model being trained on scratch over Power9 platform and via transfer learning on vgg16 ang vgg19, sample model in inference could be seen below

https://github.com/arshad2101/retinopathy/blob/master/Retinopathy.ipynb

https://openpowerfoundation.org/blogs/ai-improve-rural-healthcare/

Further working on Medical Image Classification model on Diabetic Retinopathy and get model ported on Nvidia Jetson Tx2 device for medical image classification on Edge devices.

 

Idea is to port a model on mobile device such as Jetson Tx2 and get that device connected to medical equipment’s such as retina image capturing, X-Ray, microscope, when doctors make any image capturing it could be inference on the device itself offline.

 

Developed a Face Recognition and object detection system using opencv Haar Cascade

 

Developed a classification model using Convolutional Neural Network, get it trained in PowerAI, deployed the model in Android device using tensorflow lite.

 

Worked on PowerAI Vision to Classify Breast Cancer images.

 

Working on Deep Learing model on for Face Expression detection using Tensorflow.

build a solution using Python Jupyter notebook to analyze the model quality of Watson

NLU,NLC and Conversation services and data quality of the services using NLP Latent Semantic Analysis technique

https://gist.github.com/arshad2101/085d4f03bb4d94505dc654fac7a4fb0a