Browsing by Author "Akter, Sanzida"

Now showing 1 - 3 of 3

A system to prevent social violence using convolutional neural network
(BRAC University, 2023-01) Akter, Sanzida; Omar, Mostafa Nayeem; Siam, Aanan Ehsan; Rahman, Fariha; Tanjib, Sadib; Rahman, Md. Khalilur
Most women face violence in public and at home, including rape, physical and emo tional abuse, mocking, and harassment. A social violence support system might allow people to seek aid from their friends, or relatives, or even request administra tive assistance. The goal here is to detect clearly and reliably the screams of the individual in the position that is in any danger, that is, if the scream arose out of dread and horror, based on a particular collection of audios. Screams elicited by dread and panic usually have a shorter length, a higher frequency, and shrill pitches, whereas screams elicited by other emotions or intentionally have a longer duration, a fixed frequency, and pitch. In this sense, if we can use scream recognition to recognize dangerous and consequential circumstances in our society and inform the appropriate individuals at the appropriate moment, we will be able to avert these issues to a degree that will benefit both society and its citizens. To assist the wider populace, we have implemented a system using Convolutional Neural Network to identify screams automatically. This model will assist us in recognizing screams and sending SOS signals or messages to suitable contacts. As a result, people who are in danger will be able to call the people from their selected contacts or general authorities who are within their reach at any time. This system will not only assist victims in avoiding danger, but it will also provide them with a sense of security. On the other hand, the general authority will be able to use this software to limit the quantity of social and domestic violence.
Assessment of the effect of Aegle marmelos on Gastrointestinal Motility in Albino Mice
(East West University, 12/18/2017) Akter, Sanzida
Bael (Aegle marmelos (Linn), family Rutacae, is also known as Bale fruit tree, is a moderate sized, slender, aromatic tree. The different parts of Bael are used for various therapeutic purposes, such as for treatment of Diabetes, Asthma, Anaemia, Fractures, Healing of Wounds, Swollen Joints, High Blood Pressure, Jaundice, Diarrhoea, Healthy Mind and Brain, Typhoid, Troubles during Pregnancy. Aegle marmelos has been used as an herbal medicine for the management of diabetes mellitus in Ayurvedic, Unani and Siddha systems of medicine in India, Bangladesh and Srilanka. The main usage of the parts of this tree is for medicinal purposes. The unripe dried fruit is astringent, digestive, stomachic and used to cure diarrhea and dysentery. Diabetes is increasingly affecting a growing number of patients and seriously reducing their quality of life. Use of conventional drugs in diabetes management is expensive, thus, unaffordable to most patients. Furthermore, most of these conventional drugs are associated with undesirable side effects. Incorporation of herbal medicine into conventional healthcare system may significantly improve the overall healthcare system. Evaluation of efficacy and safety by scientific method is necessary to validate herbal medicine utilization, in most cases even where efficacy of the plants has been established the standard dosage required to bring about healing is not clear. The intraperitoneal route of herbal extract administration was found to be more effective than the oral route. Our present studies were focused on the probable effect of Aegle marmelos on Gastrointestinal Motility.
Recognition of Bangladeshi Sign Language (BdSL) Words using Deep Convolutional Neural Networks (DCNNs)
(Ital Publications, 2023-12-01) Haque, Aminul; Pulok, Rishad Amin; Rahman, Md. Mizanur; Akter, Sanzida; Khan, Nusrat; Haque, Shamsul
In a world where effective communication is fundamental, individuals who are Deaf and Dumb (D&D) often face unique challenges due to their primary mode of communication—sign language. Despite the interpreters' invaluable roles, their lack of availability causes communication difficulties for the D&D individuals. This study explores whether the field of Human-Computer Interaction (HCI) could be a potential solution. The primary objective is to assist D&D individuals with computer applications that could act as mediators to bridge the communication gap between them and the wider hearing population. To ensure their independent communication, we propose an automated system that could detect specific Bangla Sign Language (BdSL) words, addressing a critical gap in the sign language detection and recognition literature. Our approach leverages deep learning and transfer learning principles to convert webcam-captured hand gestures into textual representations in real-time. The model's development and assessment rest upon 992 images created by the authors, categorized into ten distinct classes representing various BdSL words. Our findings show the DenseNet201 and ResNet50-V2 models achieve promising training and testing accuracies of 99% and 93%, respectively.