Research on image text recognition based on canny edge detection algorithm and k-means algorithm
F Wu, C Zhu, J Xu, MW Bhatt, A Sharma - International Journal of System …, 2022 - Springer
The latest research in the field of recognition of image characters has led to various
developments in the modern technological works for the improvement of recognition rate …
developments in the modern technological works for the improvement of recognition rate …
The multimodal sentiment analysis in car reviews (muse-car) dataset: Collection, insights and improvements
Truly real-life data presents a strong, but exciting challenge for sentiment and emotion
research. The high variety of possible 'in-the-wild'properties makes large datasets such as …
research. The high variety of possible 'in-the-wild'properties makes large datasets such as …
[PDF][PDF] Reducing the annotation burden: deep learning for optical character recognition using less manual annotations
C Bartz - 2022 - publishup.uni-potsdam.de
Text is a ubiquitous entity in our world and daily life. We encounter it nearly everywhere in
shops, on the street, or in our flats. Nowadays, more and more text is contained in digital …
shops, on the street, or in our flats. Nowadays, more and more text is contained in digital …
Extracting textual overlays from social media videos using neural networks
Textual overlays are often used in social media videos as people who watch them without
the sound would otherwise miss essential information conveyed in the audio stream. This is …
the sound would otherwise miss essential information conveyed in the audio stream. This is …
Community Learning Analytics Support for Audio-Visual Web-Based Learning Contents: The CIDRE Framework
An abundance of audio-visual (AV) learning content is available on the Web, eg in massive
open online courses (MOOC), in Webinars and on AV streaming platforms. However, the …
open online courses (MOOC), in Webinars and on AV streaming platforms. However, the …