1. Domain and Context
Currently autonomous driving is an area of interest and research. We would like to solve a problem in this domain and for solving similar instance segmentation problems.
Car manufacturers are widely using Artificial Intelligence to offer variety of smart solutions in the automotive industry. One such solution is Autonomous/Self driving cars. However, to develop car which is completely autonomous in nature, it should be able to identify varied range of objects just like humans are able to recognize in real life.
Our team's approach is to use Advance Computer Vision techniques such as Semantic segmentation and instance segmentation to help AI enabled cars to identify different objects while driving and respond accordingly.
Our target is to help AI enabled cars understand and differentiate between different objects like car, motorcycle, bicycle, pedestrian, truck, bus, and tricycle.

2. Problem Statement

Our team's approach is to use Advance Computer Vision techniques such as Semantic segmentation and instance segmentation to help AI enabled cars to identify different objects while driving and respond accordingly.
The scope of this project includes identifying each object instance of each pixel for every known object within an image, which are car, motorcycle, bicycle, pedestrian, truck, bus, and tricycle. Detect these objects efficiently by performing detection and segmentation for self-driving cars. While driving, it is highly important to be able to quickly detect objects, the difference between them whether it is a vehicle, person …etc.

• Developed command line interface for MCTP NVMe-MI protocol using Python; it’s a framework responsible to communicate with hardware device using MCTP request-response packets via promira/aavark exerciser.
• Developed automation scripts in PERL to perform testing for nvme protocol based firmware for Toshiba SSD.
• Understanding of executing RPC based procedure calls.
• Developed sanity scripts; executed every time firmware is released to Japan Team.
• Implemented automation testing for features such as set, get logpage, security send receive and device self-test for NVMe protocol.

• Understanding of openstack environment and hands on experience of implementing POC for cinder plugin using LVM.
• Developed python scripts for automating REST API testing for web portal.
• Hands on experience of working with ROBOT framework to automate test cases and executing python scripts.
• Currently working on VMware SRM project for fixing bugs in SRA.

• Developed compression utility using C; that interacts with compression driver on block devices via system calls/ioctls.
• Developed snapshot management utility using python over REST API; that allows SMCLI tool to create snapshot volume for logical volumes in use.
• Implemented support instant snapshot, scheduled snapshot and periodic snapshot on logical volume for user.
• Snapshot for logical volume was created using LVM and related commands.
• Developed export feature for logical volume over NFS.
• Implemented support for exporting snapshot over NFS to the client.
• Implemented statistics feature for monitoring various statistics for storage array, also used rrdtool to save statistical information in database.
• Defect fixing and self-review of code.
• Implemented verification logic for various features in automation framework.

• Followed agile process model for entire project.
• Understanding of Functional Requirements and resolving related doubts by means of query portals or video conferencing with client.

• Configured and worked on Syskaku Board (Hardware used to implement and    test application before deploying it on camcorder).

• Implemented hardware key mapping and handling mechanism.
• Implemented various views (Shooting Functions) and widgets for camera.
• Bug Fixing and Review.

• Knowledge transfer of resource and view/widget development activities amongst team.
• Follow design document for implementation and perform unit testing of each modules.
• Track status of implementation, perform self review and deliver modules to client on time.

• Releasing tested code on GIT in order to deliver source code to client.
• Working experience on GIT (Version Control System).