Praveen Eppa

Title: Praveen Eppa: Innovator in Deep Neural Network Inference

Introduction

Praveen Eppa is a notable inventor based in Hyderabad, Telangana, India. He has made significant contributions to the field of deep learning and neural networks. With a total of two patents to his name, Eppa is recognized for his innovative approaches to image processing and quantization techniques.

Latest Patents

Eppa's latest patents include a method for dynamic quantization for deep neural network inference systems. This method involves dynamically quantizing feature maps of a received image. The process includes convolving an image based on predicted maximum and minimum values, trained kernel weights, and the image data. The input data is quantized based on these predicted values, and the output of the convolution is computed into an accumulator and re-quantized. The re-quantized value is then output to external memory. The predicted minimum and maximum values are calculated based on previous values using a weighted average or a predetermined formula. Initial values are established using known quantization methods, which are utilized to initialize the predicted values in the quantization process.

Career Highlights

Praveen Eppa is currently employed at Texas Instruments Corporation, where he continues to develop innovative solutions in the field of technology. His work focuses on enhancing the efficiency and accuracy of deep neural networks, contributing to advancements in artificial intelligence.

Collaborations

Eppa collaborates with talented coworkers, including Kumar Arrakutti Desappan and Manu Mathew. Together, they work on various projects that push the boundaries of technology and innovation.

Conclusion

Praveen Eppa is a distinguished inventor whose work in deep neural network inference has the potential to revolutionize image processing techniques. His contributions to the field are significant, and his ongoing work at Texas Instruments Corporation promises to yield further advancements in technology.