Niranjan Koduri

Title: Inventor Spotlight: Niranjan Koduri

Introduction

Niranjan Koduri is a prolific inventor based in Pleasanton, California, with a total of 7 patents to his name. Known for his innovative contributions in the field of image processing, his work has significantly advanced the capabilities of image classification and deep learning technologies.

Latest Patents

Among his latest patents, one of the most notable inventions is related to image classification and accelerated classification training using deep learning image fingerprinting models and indexed embeddings. This groundbreaking technology generates image fingerprints (embeddings) using an image fingerprinting model, which are then indexed using an approximate nearest neighbors (ANN) model. This allows the system to identify the most similar fingerprint based on a subject embedding, even when the image is distorted, rotated, or cropped. The process involves clustering the indexed embeddings and labeling them, enabling users to contribute just a few images to create a labeled cluster.

Another significant patent includes an automated system for image classifier testing and outlier detection. This patent describes a method for generating an image classifier that can be tested almost instantaneously. The system employs image embeddings that are indexed and clustered, providing users with real-time feedback on the classifier's performance. Through Leave One Out Cross Validation (LOOCV), users receive near-instantaneous quality indications, as well as indications of outliers based on distance to the centroid calculations.

Career Highlights

Niranjan has made substantial contributions during his career at several esteemed companies. He has held positions at Netskope, Inc. and Digital Guardian LLC. His work in these organizations has focused on developing advanced technological solutions that emphasize security and efficiency in image processing.

Collaborations

Throughout his career, Niranjan has collaborated with various talented individuals, including Emanoel Daryoush and Jason B Bryslawskyj. These collaborations have fostered an environment of innovation and led to the successful development of new technologies in the realm of image classification and machine learning.

Conclusion

Niranjan Koduri continues to be a notable figure in the field of innovation and technology. His patents reflect his dedication to advancing image processing techniques, making significant contributions that enhance the capabilities of image classification and automated testing. As he continues to push the boundaries of technology, his work will undoubtedly inspire future innovations in the industry.