Document Library
Reference architectures, white papers, and solutions briefs to help build and enhance your network infrastructure, at any level of deployment.
Improving Performance of Optical Character Recognition with PaddleOCR using Intel® Distribution of OpenVINO™ Toolkit White Paper
Last Updated: Mar 24, 2025
The Optical Character Recognition (OCR) involves two stages: text detection and text recognition, to decode the text in the input image. This document presents the steps for optimizing the performance of OCR with the English version of PaddleOCRv3 (en_PP-OCRv3) model using Intel® Distribution of OpenVINO™ Toolkit. The improved accuracy of OCR is demonstrated with the real-time use case of identifying the speaker’s name in videoconferencing. Audio transcription is the process of converting the speaker’s audio into text in videoconferencing. On the other hand, diarization is the process of partitioning an audio stream containing human speech into homogeneous segments according to the identity of each speaker.
Related Content
