Embedded Files
(Volume: 2, Issue: 1)
Modern technology supports automatic disease detection in Oryza sativa
Modern technology supports automatic disease detection in Oryza sativa
Oryza sativa, commonly known as rice, is one among the staple crops that nourish almost the entire population around the globe. Besides its key role in getting rid of starvation, rice also adds to a country’s economical growth. However, the changing climatic conditions, which result in severe droughts or heavy rains, and the emergence of numerous diseases in crops make the farmers to hope for heavy loss, rather than good returns. Bacterial Leaf Blight (BLB) is a very serious bacterial disease in rice crops, caused by Xanthomonas oryzae pv. oryzae. So, when can one find the onset of this disease? It is when the leaves turn out wet and yellow to cause lesions or when the seeds start to wilt at an earlier stage, questioning both the quality as well as the quantity of the cultivated rice. Do these lesions in the leaf of rice plants need utmost attention? Obviously yes, since they act as the home for numerous disease-causing pathogens, fungi and similar other destructive microorganisms. As the rice fields are denser and wet for most of the time, the BLB disease spread during strong winds and rainy weather is not at all a wonder. Hence, prior BLB malady detection could prevent farmers from experiencing huge loss. The question now is “Can the agricultural experts help in this regard?” Certainly…They can identify the disease and provide measures to alleviate and control the disease. Yet, the detection accuracy and the large analysing period triggers the need for an automatic detection system, which may or may not deploy the expert’s knowledge. Two researchers from Tamil Nadu, India, have classified the rice crops as BLB-affected, normal and other disease-affected using the image of rice crops in a novel Deep Convolutional Neural Network architecture. The researchers in their article in Advances In Engineering Software, vol.174, Issue: C, Dec 2022, rendered 97.5% disease-finding accuracy. Eventually, they also suggested that the incorporation of IoT-related observing images could well-aid the future research on automatic rice disease detection. Further to mention, in contrast to rice disease detection, research on the automatic finding of manure proportions and humidity levels to prevent the BLB occurrence can be yet another research direction for the future.
Image Courtesy: Rice Knowledge Bank
A combined mHealth and cloud-based tech to aid oral cancer detection
A combined mHealth and cloud-based tech to aid oral cancer detection
The public’s ignorance to the incessant warning about the deleterious effects of alcohol, areca nut and tobacco consumption has ended up in a worse scenario in 2020 - ‘Oral cancer, the thirteenth common cancer around the world’. This cancer type, developing on and around the tongue, the tissue linings of the gums as well as the mouth and the throat region behind the mouth can be roughly diagnosed from about 54540 new cases in 2023, expecting to take away the lives of 11580, declares the American Cancer Society. So, early diagnosis is the sole process that can help in this regard, leading to early recovery. Normally, the diagnosis demands a dental expert’s observation of the cancer-affected area using the essential equipment. But, how do the people living in remote areas who comparatively contribute a larger percentage to oral cancer, acquire a dental examination or an oral cancer diagnosis? Does the people of this busy world always have time to visit the expert and get medical assistance? It is here the mobile health (mHealth) and cloud-based technology comes into play. As with mHealth, people could obtain their healthcare services and health research or outcomes through mobile/ wireless technology. In contrast, the cloud-based technology allows people or the expert to furnish or acquire their medical data and services at any time, without the complete active intervention of the cloud users. Keeping the technological benefits in mind, Savita Shetty and Annapurna P Patil, the researchers from Visvesvaraya Technological University, Bangalore, have modelled a two-tier oral cancer detection system. They state that the medical image data of the user from distinct places can be stored in respective clouds, which can then be processed for automatic cancer detection, making the outcomes and medical services to be rendered to the users through mHealth devices. The 92.17% accuracy from their novel approach is a great boon for the present world. However, the research thirst towards attaining improved oral cancer detection with increased data security and universal access to mHealth and cloud computing platforms is always encouraged…
Image courtesy: www.freepik.com
Text-to Image Synthesis for Image and Graphic Designing
Amazed to see your brain’s visualization on a computer or a mobile screen by just uttering the relevant text? It is nothing but the sequel to Text-to-Image Synthesis- A computer vision task encompassing Artificial Intelligence (AI) and machine learning approaches, which is believed to master the future of Brain-Computer Interface (BCI), graphic designing, computer-aided design, image editing and many more emerging technologies. By text-to-image synthesis, one can use the AI techniques to process his/her natural language for creating more realistic picture encodings that match the spoken text. So, who can effectively achieve this job? The answer at present might be the Deep Learning approaches. To be specific, the Generative Adversarial Networks (GAN). Usually, the GAN has a generator and a discriminator, working in an adversarial sense to produce highly photo-realistic images in an image-to-image translation process. The GANs in text-to -image synthesis task too works in the same way, except that the training features in both the networks include several textual descriptions, even for a single image.
The research published in Advances in Engineering Software, vol. 173, too supports this notion of text-to-image synthesis using optimized GANs. “The proposed model can produce complex images on various datasets with high fidelity features. However, the image quality rapidly declines as the text includes several scenario settings and occurrences. In the future direction, this work will concentrate on producing many high-resolution photos from a single semantic layout and leveraging knowledge graphs to infer similar semantic layouts”, the researchers say.
Can research on text-to-image synthesis be beneficial for future? Absolutely. Creating an art with just a few imaginary words for computer-aided design or graphic designing is beyond imagination. The teaching mode can be transformed such that the students can develop their intellect with synthesized images from unexplainable textual facts. Further, the forensic sector could use this tech, while querying about the crime. However, all the benefits fall in the hands of users, who make it fraudulent or not!!!
Image courtesy: www.pixabay.com
Intrusion Detection- A big challenge for the big data in cloud
Cloud computing and big data handling, without compromising security, is the principal need of any internet-based organization or service in this digitally-busy computerized era. The clouds, generally referring to a group of network-based computing resources and servers performing distinctive operations, is always stored with massive data in structured, unstructured or semi-structured forms. The simple “pay as you use” nature of cloud computing, though allowing access to excessive data and server usage, has become a spot of insecurity nowadays!!! The unauthorized cloud users or intruders are the main cause and this insecure state has given rise to “Intrusion Detection Systems” (IDS). Denoting a device or a software system, the IDS’s functionality in the cloud is “Beware user!!! A malicious or policy-violated occurrence is detected in the cloud”. The IDS examines the network traffic or the activities of the host device or a combination of both to make this alert. However, the unmanageable high-volume and high-speed data processing ability of the big data cloud networks deceive the IDS about the malware presence, even if numerous IDS are installed without considering the cost associated. The security task is now entirely at risk!!! So, how research can support to overcome this insecurity? In three ways…One is to put forth approaches that allow the cloud data or resources to be manageable and clustered, aiding the IDS to effectively spot the intruder. The second way is to decide on the optimal number of IDS or its optimal placement in the cloud to handle the massive data. The final way is the utilization of optimal cloud features to detect the intrusion, provided that the network complexity or cost fall within limits. Two researchers each from India as well as Peru and one from Italy have paid attention to the selection of optimal big data features for achieving intrusion detection in the cloud environment. On using the datasets like, MQTT-IOT-IDS2020 and the Apache Web Server dataset, the researchers have aimed an intrusion-free cloud storage and usage with better accuracy, kindling minds to perform further research in this domain.
Image Courtesy: www.vecteezy.com
Energy-aware resource allocation for reliable avionics systems
Can the unused organizational resources be used in a sustainable way? Can the proper as well as the optimal utilization and distribution of such resources cut down the cost associated with additional resource requirements? Does such optimal consumption of resources improve the product reliability? If unsure, the answer is “Definitely possible through resource planning and allocation, provided that the organizational goals are not being violated”. Resource planning and allocation plays a significant part in an organization’s success, owing to its ability to extract the maximal use of resources at unused times, without hindering the chain of processes within the organization. The rapidly progressing aviation industry is also not exempted from using resource allocation strategies, especially due to the energy efficiency in demand. Lots of interconnected electronic hardware devices, which group to form the avionics system, are the resources demanding high energy efficiency… Be it during their energy consumption or their energy release period. Optimally allocating the available avionic hardware can help reduce the cost of additional devices performing identical operation, the greenhouse gas emissions and the heat dissipation that questions the reliability of the final avionics’ product. Currently, software-based resource allocation has gained worldwide interest because of its ease in establishing task management and control. The China-based research on Computers and Electrical Engineering, Elsevier, vol. 105, presents one such software-based resource allocation module using improved ant colony optimization algorithm. Providing superior results than the state-of-the-art methods in minimizing the makespan and energy consumption of avionic resources, the researchers foresee a reliable Quality of Service (QoS) with consideration on additional optimization objectives.
Image Courtesy: www.pixabay.com
Page updated
Google Sites
Report abuse