Utsa Sen Email & Phone Number

Stuttgart, Baden-Württemberg, Germany

Topic: Unsupervised Domain Adaptation for 2D Object Detection- Generate Taget domain dataset using stable diffusion model- Incorporate Instance copying from source dataset to Target domain dataset and train knowledge distillation based unsupervised domain adaptive model (Teacher Student Strategy) in Detectron 2 (Meta) Framework- Integrate Yolov7 into the Detectron 2 architecture. Supervisor: George Basem Fouad EskanderExaminer: Prof Dr Ing. Bin YangDepartment: Institut für Signalverarbeitung und Systemtheorie (ISS)

Stuttgart, Baden-Württemberg, Germany

-Introduction to CAN-Real-time programming with Ada-Introduction to Microcontroller Programming-PLC – Programmable Logic Controllers-Rapid Prototyping with ASCET-SD-Test automation with Hardware-in-the-Loop

Stuttgart, Baden-Württemberg, Germany

Grade: 1.3(Sehr Gut)/1.0(Best Grade)Topic: Improving Domain Generalization of Image to Image translation models using Deep Neural Network-Main application of the project was to develop techniques such that autonomousvehicles can be able to segment road scenes reasonably well on foggy, noisy real worldimage data which it had not seen during training.-Trained the developed model on Cityscapes(free from real-world fog, noise), validated on a different domain -GTA, evaluated on foggy cityscapes datasets.-Researched ways to make the model independent of the input data's style and texture by implementing various techniques.-Developed a user-selectable channel-wise Batch/Instance Normalization block whichimproved the overall generalization on the unseen domain from mIoU 45% to 46.94%-Finally improved the generalizability of a standard state-of-the-art image segmentationmodel RobustNetDepartment: Institut für Signalverarbeitung und Systemtheorie (ISS)Supervisor : Chenming JiangExaminer: Prof Dr Ing. Bin Yang*Tags: DeepLabV3+, IBN-Net, PSPNet, Semantic Segmentation, Pytorch, Computer Vision

Stuttgart, Baden-Württemberg, Germany

-Worked in a multidisciplinary team on labeling differenttypes of Infrastructures, defects and visual damages inconstruction elements for a neural network (AI system).-Worked on Damage optimization and study of neural networkalgorithms applicable.-Gained knowledge on Deep Learning concepts.

Pune Area, India

Field Engineering (ROW - Rest Of the World Team)1. Responsible for designing HVAC systems based on customer inputs2. Responsible for developing engineering solutions for the HVAC systems, understandingcustomer requirements, proposing control logics, preparing a complete engineeringpackage.3. Responsible for Hardware & Software designing of HVAC systems.4. On-Time Delivery and Error Free deliverables to the various JCI branch customersglobally.5. Responsible for ensuring process compliance and adherence within the team.

Pune Area, India

1. Basic Hardware Engineering for the Standard Systems, Damper / Room/ ValveSchedule.2. Logic Program Designing and Graphics Design for the user interface.3. Preparing Metasys System Configuration Tool (SCT) Database and Graphics for BAS.4. Development and up-gradation of Graphics templates, Advance graphics symbol library.5. Job or project review and man-hour estimation 6. Hands-on training encompassingdynamic aspects of the Building Management systems.

Kolkata, West Bengal, India

Automatic Car Parking based Indication System.Components Used:-Microcontroller (AT89S52/AT89C51)-Photo Diode-LDR-LED-DC Motor

Kolkata

Home Security System using 8051 Microcontroller

Master Of Science - Ms, Electrical Engineering

Activities and Societies: Photography Club Major: Smart Information Processing Courses Attended: Computer Vision, Machine Learning, Deep.

Bachelor Of Technology (Btech), Electronics And Instrumentation Engineering, 1.8/1.0 (Topped In The Class In Terms Of Overall Gpa)

Courses Attended: Analog Electronics Circuits Digital Electronics Circuits Microprocessor and Microcontroller Circuit Theory and Networks.

Science

Courses Attended : Math, Physics, Chemistry, Biology, English Literature, English Language, Bengali, EVS