Face recognition using DNN (Internship Project Summer 2019)

– Internship applications are now closed .

– Thank you for such an overwhelming response.
Sorting and selection of such high quality profiles took a bit longer than expected for 500+ applicants.
– We want to support as many students as possible.

Keeping this in mind, beyond the on-site internship we have added a remote “Learners” group that can learn and contribute by following the main group’s work, online.

– Please scroll-down to see the list of selected students for next round.

To intern with us on this project we have some pre requisites :
1)Must be sincere and committed .
2)Must be self motivated to finish goals
3)Must have completed at least one personal technical project , regardless of stream .


For sincere candidates we provide following Free training :

TRAINING (2 stages)
Stage 1
Python: Basics.
OpenCV: Basics, live video analysis, image operation , framing.
Python data science libraries: Numpy, Pandas.
ML basics, DNN basics.

Tools training : Phabricator , Git , project management .

Stage 2

Tensorflow low level API understanding.
Complete training on Face Recognition using Tensorflow
>Understanding general Facial recognition theory.
>Model training: triplet loss
>Face detection
>Finding landmarks
>Generating embeddings
>Recognition using Euclidian distance 

Porting model on edge(embedded device )
Basic model optimisation theory
Neural Network  compression
Introduction to BNN


Live Project Goal:

To build end to end scalable real-time facial recogntion system on edge device


Tasks :

>Implement Single sample per person recognition(SSPP)

>Research with goal to improve current model or find better models.

>Implement facenet architecture on lightweight neural net backbone

>Implement face tracking during inference

>Implement best classification algortithm on trained embeddings for large scale data

>Implement other distance metrics to compare the embeddings

Other Project goals :

>Studying and implement binary neural networks, Ternary Weight Networks, DorReFa-Ner
>Implement embedded BNN papers using C++
>Making FR application using BNN


We now plan to support more than one team due to the enthusiastic response received.

Students are allowed to take a project goal of their choice and expect continued support from us post internship . We may sponsor their idea if it has startup potential since the primary mission of this community is to create innovative technical startups 

Selected Students for Next Round

Gourav BarnwalIIT Bhubaneswar
Rushikesh WayalIIT Kharagpur
Kanishk AgarwalIIT Kharagpur
Shishir GoyalIIT Kharagpur
Tanya SnehIIT Kharagpur
Aditya PrakashIIT kharagpur
Pranay MandarIIT Madras
Abhishek KumarIIT Roorkee
Mahesh MedamIndian Institute Of Technology- Kharagpur.
Shah Nawaz AlamIndian Institute Of Technology- Kharagpur.
S Uday BhaskarIndian Institute Of Technology- Kharagpur.
Jigyasa SrivastavaIndian Institute of Technology, Kharagpur
Nisarg ShahNational Institute of technology , karnataka
Aryaman SriramVellore Institute of Tech
Prakhar PriyeshKIIT Bhubaneshwar
Arnav KarmarkarPICT
Ankita KulkariPICT College
Pranav GuptaMIT WPU
Purvi ModiModern Education Society’s College of Engineering , Pune
Shivam SharmaIIT Bhubaneswar

List of students who will be in learners group. Status subject to change on further rounds.

Khush PanchalIIT Kharagpur
Yash KhivasaraNational Institute of Technology Karnataka
Umesh ShahdadpuriIndian Institute of Technology Delhi
Stephen masihKiit Bhubaneswar
Abhishek KaushalIIT kharagpur
Akash Pratap SinghIndian Institute of Technology Bombay
Omkar IngaleIndian Institute of Technology Bombay
Nikhilesh ReddyIndian institute of information technology Pune
Abhinav KumarIndian Institute of Information Technology, Pune
Munish BhardwajIndian Institute of Information Technology, Pune
Susmita KaryakarteMMCOE
Aayush UpadhyaySRM Institute of Science and Technology
Apoorva PiseCOEP
Nachiket Rajendrasing RajputCollege Of Engineering Pune
Chaudhari Tejas SharadCollege of engineering, pune
Pakhi DahiyaIIT Delhi
Murali MridhulaUniversity of Surrey

Those who aren’t selected for the internship. We thank them for applying and wish them luck for our future internships.


All those who are selected and are in learners group, congratulations to all of you.Please fill this google form for next round.

Click Here


We have added preliminary overview of internship for now. More details to be added soon. Subject to change. Check out here.


©[2018]  No content is copyrighted ,use freely.IoTIoT.in is part of Nikhil Bhaskaran Foundation

Log in with your credentials

Forgot your details?