Computer Vision Engineer - Image Processing

PARLE GLOBAL TECHNOLOGIES PVT. LTD
Maharashtra
5 - 8 Years

Posted on: 17/05/2025

Job Description

Key Responsibilities :

- Design and develop CV algorithms to analyze RGB and 3D video streams for human and infant behavior monitoring

- Translate scientific concepts such as optical flow, signal extraction, and motion tracking into efficient, production-ready code

- Lead and execute data collection campaigns, including annotated dataset creation under varied lighting and real-world conditions

- Build CV pipelines in Python using OpenCV, classical vision techniques, and deep learning models for movement, posture, and sleep detection

- Cross-compile and deploy computer vision modules to ARM-based SoCs, optimizing for CPU/GPU utilization and low power consumption

- Conduct extensive data analysis to identify vision patterns and improve model accuracy and robustness

- Collaborate with hardware, firmware, and product teams to integrate CV modules into a production-grade monitoring system

Required Skills & Qualifications :

- 5+ years of experience in Computer Vision, Image Processing, or Signal Processing roles

- Strong hands-on expertise with Python and OpenCV for image filtering, motion detection, SVD, PCA, and optical flow techniques

- Experience working with deep learning frameworks such as TensorFlow or PyTorch

- Familiarity with model quantization and optimization for edge inference

- Solid understanding of linear algebra, signal extraction, and DSP techniques

- Experience in real-time system design on embedded Linux platforms

- Ability to handle large-scale annotated video datasets and derive meaningful insights

Preferred Qualifications (Good to Have) :

- Prior experience in baby monitoring, health analytics, or related real-time tracking systems

- Experience with cross-compilation, deployment, and performance tuning on ARM-based SoCs

- Understanding of embedded resource constraints and low-power design principles


info-icon

Did you find something suspicious?