Looking for a job

Shaunak Atul Kanikar

@shaunakatulkanikar

Machine Learning Engineer at GFaI who developed a PatchCore textile inspection demonstrator that detected anomalies in over 98% of affected frames.

Germany

Message

What I'm looking for

I’m looking to bring applied deep learning to human motion capture and pose estimation.

At Society for the Advancement of Applied Computer Science (GFaI), I developed a camera-based PatchCore demonstrator for detecting anomalies in complex-patterned textiles. Tested on textiles from real SME use cases, it detected anomalies in over 98% of frames that contained them.

At GFaI, I also designed an end-to-end pipeline to predict polymer creep behavior from partially missing industrial data, achieving ~2% MAPE on held-out data. For acoustic quality inspection, I used one-class anomaly detection to improve classification of defective workpieces.

At Simi Reality Motion Systems GmbH, I developed a query-based deep learning architecture for continuous landmark detection in driver pose estimation and built its end-to-end pipeline in TensorFlow. I’m looking to bring applied deep learning to human motion capture and pose estimation.

Experience

Work history, roles, and key accomplishments

Current

Machine Learning Engineer

Current

Society for the Advancement of Applied Computer Science (GFa

Oct 2024 - Present (2 years)

Developed a PatchCore-based anomaly detection demonstrator for textiles, detecting anomalies in over 98% of anomalous frames. Designed end-to-end ML pipelines for polymer creep prediction and acoustic quality inspection, achieving ~2% MAPE on creep regression.

Python PyTorch TensorFlow scikit learn XGBoost LightGBM CatBoost Pandas NumPy OpenCV Docker Git MLFlow W&B

Master's Thesis in Computer Vision

Simi Reality Motion Systems GmbH

Oct 2023 - Apr 2024 (6 months)

Developed a query-based deep learning architecture for continuous landmark detection in driver pose estimation, using a ResNet-50 encoder and SMPL-based query encoder. Built an end-to-end pipeline with TensorFlow.

Python TensorFlow Deep Learning Computer Vision ResNet SMPL MLP

Computer Vision Intern

Simi Reality Motion Systems GmbH

May 2023 - Aug 2023 (3 months)

Evaluated Time-of-Flight cameras for depth accuracy, reflectance sensitivity, and temporal stability. Designed and executed lab experiments and developed Python scripts for data acquisition and analysis.

Python Computer Vision Time Of Flight Cameras Sensor Performance Evaluation Data Analysis APIs and SDKs

Education

Degrees, certifications, and relevant coursework

RWTH Aachen University

Master of Science, Robotic Systems Engineering

2020 - 2024

Grade: 1.7

Master of Science in Robotic Systems Engineering from RWTH Aachen University, with thesis on continuous landmark detection for driver pose estimation.