Redirigiendo al acceso original de articulo en 15 segundos...
Inicio  /  Future Internet  /  Vol: 14 Par: 3 (2022)  /  Artículo
ARTÍCULO
TITULO

Survey on Videos Data Augmentation for Deep Learning Models

Nino Cauli and Diego Reforgiato Recupero    

Resumen

In most Computer Vision applications, Deep Learning models achieve state-of-the-art performances. One drawback of Deep Learning is the large amount of data needed to train the models. Unfortunately, in many applications, data are difficult or expensive to collect. Data augmentation can alleviate the problem, generating new data from a smaller initial dataset. Geometric and color space image augmentation methods can increase accuracy of Deep Learning models but are often not enough. More advanced solutions are Domain Randomization methods or the use of simulation to artificially generate the missing data. Data augmentation algorithms are usually specifically designed for single images. Most recently, Deep Learning models have been applied to the analysis of video sequences. The aim of this paper is to perform an exhaustive study of the novel techniques of video data augmentation for Deep Learning models and to point out the future directions of the research on this topic.

 Artículos similares

       
 
Joël Meissonnier, Virginie Dejoux     Pág. 4403 - 4409
The French Act of 11 February 2005 on equal rights and opportunities, participation and citizenship of people with disabilities highlights the need ?for people with disabilities or reduced mobility? to have access to the complete mobility chain of the tr... ver más

 
Odette Laneuville and Dorota Sikora    
To improve the learning of basic concepts in molecular biology of an undergraduate science class, a pedagogical tool was developed, consisting of learning objectives listed at the end of each lecture and answers to those objectives made available as vide... ver más
Revista: Future Internet