Senior Data Scientist (Computer Vision)

Key Skills

Java Programming, Python Programming, SQL, Deep Learning, Keras software library, TensorFlow, Caffe, OpenCV

Job Description

• Research and Develop Innovative Use Cases, Solutions and Quantitative Models in Video and Image Recognition and Signal Processing for Hitachi’s cross-industry business (e.g., Energy, Industry, Mobility, Smart Life and Financial Services).
• Design, Implement and Demonstrate Proof-of-Concept and Working Proto-types for Hitachi and its clients.
• Provide R&D support to Hitachi business units and group companies to productize research prototypes.
• Explore emerging tools, techniques, and technologies, and work with academia for cutting-edge solutions.
• Collaborate with cross-functional teams and eco-system partners for mutual business benefit.
• Generate eminence via patents, publications, thought leadership, invited speakerships and conspicuous presence in forums of repute. 
• Add value to self through continuous learning and knowledge acquisition.
• Give back learnings to colleagues and communities.
• Mentor colleagues for growth and success.

Mandatory Requirements:

Academic Qualification:

Bachelor’s degree with STEM background (Science, Technology, Engineering and Management) with strong quantitative flavour.

Strong Fundamentals:

The candidate is required to build quantitative models from first principles and hence need to have excellent understanding of basics in mathematics and statistics (e.g., differential equations, linear algebra, matrix, combinatorics, probability, Bayesian statistics, eigen vectors, Markov models, Fourier analysis).

Core Expertise:

The candidate is expected to specialize in Video and Image Signal Processing. Solid understanding of Video and Image Recognition is a pre-requisite. It is good to have hands-on model implementation skills for any one of the below areas.

Image segmentation:

• Thresholding, edge detection and linking, region growing.
• Deformable shapes, active contours, etc.

Image representation:

• Basic descriptors: area, minor axis length, major axis length, normalized axis ratio, eccentricity, Fourier descriptors, shape numbers, etc.
• Binary descriptors: ORB, BRISK, LBP, etc.
• Advanced descriptors: HoG, SIFT/SURF, GLOH, etc.
• Object detection/classification (using classical image processing & Deep Learning)
• Image/background modeling, change/anomaly detection using image processing.
• Technology stack (e.g., OpenCV, Pillow, torchvision, TensorFlow, Caffe, Python/Keras).

• Research and Develop Innovative Use Cases, Solutions and Quantitative Models in Video and Image Recognition and Signal Processing for Hitachi’s cross-industry business (e.g., Energy, Industry, Mobility, Smart Life and Financial Services).
• Design, Implement and Demonstrate Proof-of-Concept and Working Proto-types for Hitachi and its clients.
• Provide R&D support to Hitachi business units and group companies to productize research prototypes.
• Explore emerging tools, techniques, and technologies, and work with academia for cutting-edge solutions.
• Collaborate with cross-functional teams and eco-system partners for mutual business benefit.
• Generate eminence via patents, publications, thought leadership, invited speakerships and conspicuous presence in forums of repute. 
• Add value to self through continuous learning and knowledge acquisition.
• Give back learnings to colleagues and communities.
• Mentor colleagues for growth and success.

Mandatory Requirements:

Academic Qualification:

Bachelor’s degree with STEM background (Science, Technology, Engineering and Management) with strong quantitative flavour.

Strong Fundamentals:

The candidate is required to build quantitative models from first principles and hence need to have excellent understanding of basics in mathematics and statistics (e.g., differential equations, linear algebra, matrix, combinatorics, probability, Bayesian statistics, eigen vectors, Markov models, Fourier analysis).

Core Expertise:

The candidate is expected to specialize in Video and Image Signal Processing. Solid understanding of Video and Image Recognition is a pre-requisite. It is good to have hands-on model implementation skills for any one of the below areas.

Image segmentation:

• Thresholding, edge detection and linking, region growing.
• Deformable shapes, active contours, etc.

Image representation:

• Basic descriptors: area, minor axis length, major axis length, normalized axis ratio, eccentricity, Fourier descriptors, shape numbers, etc.
• Binary descriptors: ORB, BRISK, LBP, etc.
• Advanced descriptors: HoG, SIFT/SURF, GLOH, etc.
• Object detection/classification (using classical image processing & Deep Learning)
• Image/background modeling, change/anomaly detection using image processing.
• Technology stack (e.g., OpenCV, Pillow, torchvision, TensorFlow, Caffe, Python/Keras).