Megh Makwana - NVIDIA | LinkedIn (original) (raw)

About

Applied Generative AI Engineering Manager working at the intersection of building…

Activity

14K followers

Experience & Education

View Megh’s full experience

See their title, tenure and more.

Licenses & Certifications

Deep Learning Workshop

NVIDIA

Issued Dec 2017 Expires Dec 2017

Publications

ICDAR 2019 May 21, 2019

Recognizing text from degraded and low resolution document images is still an open challenge in the vision community. Existing text recognition systems require a certain resolution and fails if the document is of low resolution or heavily degraded or noisy. This paper presents an end-to-end trainable deep-learning based framework for joint optimization of document enhancement and recognition. We are using a generative adversarial network (GAN) based framework to perform image denoising followed…
Recognizing text from degraded and low resolution document images is still an open challenge in the vision community. Existing text recognition systems require a certain resolution and fails if the document is of low resolution or heavily degraded or noisy. This paper presents an end-to-end trainable deep-learning based framework for joint optimization of document enhancement and recognition. We are using a generative adversarial network (GAN) based framework to perform image denoising followed by deep back projection network (DBPN) for super-resolution and use these super-resolved features to train a bidirectional long short term memory (BLSTM) with Connectionist Temporal Classification (CTC) for recognition of textual sequences. The entire network is end-to-end trainable and we obtain improved results than state-of-the-art for both the image enhancement and document recognition tasks. We demonstrate results on both printed and handwritten degraded document datasets to show the generalization capability of our proposed robust framework
See publication

Courses

-

-

-

Projects

Apr 2018 - Present

Building a hybrid recommendation system using collaborative and content based filtering using various machine learning and deep learning algorithms

Jan 2018 - Mar 2018

A convolutional neural network (CNN) is proposed to learn multiple useful feature representations for a classification from low level (raw pixels) to high level (object). Convolutional kernels are initialized by the learned filter kernels that come from sparse auto-encoders. Unlike some traditional methods, which divide the feature abstracting and classifier training into two separated
processes, a discriminative feature vector and a single multi-class classifier of softmax regression are…
A convolutional neural network (CNN) is proposed to learn multiple useful feature representations for a classification from low level (raw pixels) to high level (object). Convolutional kernels are initialized by the learned filter kernels that come from sparse auto-encoders. Unlike some traditional methods, which divide the feature abstracting and classifier training into two separated
processes, a discriminative feature vector and a single multi-class classifier of softmax regression are learned simultaneously during the training process. Based on the learned high-quality feature representation, the classification can be efficiently per- formed. A real-world case of surface defects on steel sheet, which evaluates the classification performance of the proposed method, is depicted in detail. The experimental results indicate that the proposed method is quite simple, effective and robustness for the classification of surface defects on hot-rolled steel sheet.

Sep 2016 - Jun 2017

Our solution is to provide the user with a real time viewing of vertically upward parking structures which enables the
user to view occupied/unoccupied parking space at the very instance, also the number of parking vehicles in the parking
space which are yet to be allocated parking space. This enables the user to check availability from their devices from
anywhere around the city via our application. To access the application authentication of user is done prior using its
service by…
Our solution is to provide the user with a real time viewing of vertically upward parking structures which enables the
user to view occupied/unoccupied parking space at the very instance, also the number of parking vehicles in the parking
space which are yet to be allocated parking space. This enables the user to check availability from their devices from
anywhere around the city via our application. To access the application authentication of user is done prior using its
service by linking their adhar card number to the database as unique identification. Due to the rapid increase of vehicles
in major cities there is going to be a need for more parking space with low area usage.
See project

Other similar profiles

Explore top content on LinkedIn

Find curated posts and insights for relevant topics all in one place.

View top content

Others named Megh Makwana in India

5 others named Megh Makwana in India are on LinkedIn

See others named Megh Makwana

Add new skills with these courses