A Survey on the State of Art Approaches for Disease Detection in Plants

Abstract

Agriculture is the main factor for economy and contributes to GDP. The growth of the economy of many countries is based on agriculture. As a result, the yield factor, quality and volume of agricultural products, play a critical role in economic development. Plant diseases and pests have become a major determinant of crop yields throughout the years, as such illnesses in plants offer a serious threat and impediment to higher yields or production in the agriculture industry. As a result, From the outset, it becomes the major duty to correctly monitor the plants, to detect diseases thoroughly, and to determine methods of controlling or monitoring these plant diseases pests in order to achieve a higher rate of production growth and minimal crop damage. Using machine vision, deep learning methods and tools for extracting and classifying features, It could be possible to build a reliable disease detection system. Numerous researchers have created and deployed various ways for detecting plant diseases and pests. The potential of these methods has been examined in this work.

A State-of-arts Review of Deep Learning Techniques for Speech Emotion Recognition

“…The use of DL techniques for emotion recognition has grown significantly in recent years as a result of developments in digital signal processing, enhancements in human-computer interfaces [20], and quick advancements in ML [21]. It is necessary to concentrate on DL techniques at present due to the growing body of research on the topic, as well as the interest in and emphasis on emotion recognition in DL [22]. The ML pipeline techniques used in these studies, which include speech feature extraction, dimensionality reduction, emotion categorization based on underlying speech features, and speech signal isolation, were mostly used to complete SER tasks.…”

“…Still, the advent of DL approaches has resulted in significant progress in this field. Improved emotion detection accuracy has resulted from the remarkable performance of CNNs and RNNs in capturing spatial and temporal relationships in emotional data [22].The three crucial steps are typically followed by the machine learning algorithms employed in these investigations: pre-processing [32], speech signal isolation, feature extraction and selection, and emotion classification from audio signals. There are some intrinsic difficulties in deducing the emotional states of speakers from their speech [33].First, it's unclear which aspects of speech are most useful for differentiating between different emotional states.…”