Browsing by Author "Zaman, Saima"

Now showing 1 - 2 of 2

Feeding practices of under-5 children in a semi-urban village in Bangladesh
(BRAC University, 2012-12) Moe, A.Z.; Sharmin, Farhana; Zaman, Saima
Objective: To explore the reported feeding practices of under-five (U-5) children in a rapidly urbanizing village in Bangladesh. Methodology: This was a study with qualitative exploratory design based on grounded theory. Researchers identified caregivers of U-5 children and shopkeepers in proximity to the village through purposive sampling. The researchers conducted five in-depth interviews, two focus group discussions, participatory rapid appraisal exercises and three informal interviews were done. For the purpose of triangulation, informal observations took place in households, nearby shops and with children on the streets. Results: The study identified five main themes: family influences, caregiver’s educational level, overall changes in time, the perception of “good” or “bad” food for health and preferences. Conclusion: Exploring these themes regarding the feeding practices of U-5 children will be helpful in understanding the health and nutritional changes of nearby semi-urban areas going through rapid development. Keywords: under-5 children, feeding practices, urbanization, Bangladesh
Road sign detection and translation in Bangla using image processing and machine learning
(BRAC University, 2019-08) Islam, Khalid Amirul; Mubin, Nadia Farha; Zaman, Saima; Alam, Md. Ashraful
For the drivers while driving, road signs play very important role but it is really challenging for the drivers' if they are not at home with the meanings of the signs. Driving is a process which is complex, continuous, and multitasking that involves driver's perceptive ability and motor movements. Road and tra c signs and vehicle information is presented in such a manner that distracts driver's attention intensely with increased mental workload leading to safety concerns. Road sign recognition and translation system is basically a system for intelligent vehicle which guides the driver to obey the tra c rules. These are the tra c rules which are represented in a small pictorial form, erected at road sides. Our proposed work represents the process of recognizing the captured sign and then translate the detected sign in Bangla. Translation of an image is done by matching multiple images. We are presenting a system for detecting and recognizing the signs displayed around us and voice synthesizing their contents with an algorithm for detection and recognition of road sign using image processing and machine learning method. We have used a template matching algorithm to match the road signs with the database after recognizing road signs it will be translated in Bangla or English according to user preference and manufactured as voice output note. The detection, recognition and speech synthesis modules each will perform their respective tasks e ectively and e ciently, and the future advancement of the currently proposed system is promising.