Search

Most of the areas of Bangladesh face a wide range of threats, ranging from terrorism and civil unrest to kidnapping, rape and murder. To reduce the impact of these threats, it is critical for the authorities to capture real-time information onwhat is happening in and around the city. Therefore, there is a growing requirement for utilizing the new and emerging technologies to make our cities safer. The development of the monitoring system in the technological field has become a blazing topic in the milieu of scholars and AI techniques have become an essential part of the future era. With the help of AI & deep learning, we can evaluate the social behavior (Action Recognition) of people, and by implementing the social credit score system, we can mark the citizen’s activities & analyze them. To detect and recognize the actions of a human we used YOLOV2 architecture and also incorporated the SQL database (SQLite) to record the social credit score. A citizen can visualize their credit score from an HTML based website.

In order to upgrade the efficiency level of multiple tracking like face, actions, characters, a deep learning method is introduced to reduce the accidents occurred in roads for carelessness and also to capture the criminals in Bangladesh. This paper presents a faster processing multiple detection method with the best possible outcome under the framework of YOLOv2 algorithm in the event of car accident, crossing foot over bridge and using the zebra crossing in Bangladesh. Different layers were added to the YOLOv2 algorithm to pass the information in various convolutional layers to detect multiple objects with actions. In this paper YOLOv2 algorithm under DarkFlow framework is used to achieve higher ratio of confidence value as the max convolutional layers reorganize the feature map so that other layers feature map can be matched with the bottom layers to achieve the expected output of the indicated events. By removing the noise from the unrelated area, the detections of the training video and test video adopt quite parallel confidence ratio.

As soon as coins or money was invented, there were people trying to make counterfeits. Counterfeit money is fake money that is produced without the permission of the state or government, usually to imitate the currency and deceive the intended recipient. In Bangladesh, this is a significant problem and the problem is becoming more and more phenomenon as the days are passing by. Today’s modern bank notes have several security features that makes easier to identify fake notes. One of the security features is the use of UV ink. Bank notes deliberately put random flecks of color scattered all over the surface of the money - which acts as a extra layer of protection against counterfeiters. We propose an automatic authentication model for identifying counterfeit money based on these random flecks of color which is visible under UV light. To obtain a benchmark result, existing object detection pre-trained models were used, followed by MobileNet, Inception, ResNet50, ResNet101, and Inception-ResNet architectures. After that, using the Region Proposal Network (RPN) method with Convolutional Neural Network (CNN) based classification the optimal model was proposed. The proposed model had a 96.3 percent accuracy. It is critical to reduce the circulation of counterfeit money in a country’s economy to stop inflation. This study will aid in the detection of counterfeit money and, hopefully, reduce its spread.

Objectives: The COVID-19 pandemic is among the most serious global threats, and it is still a significant concern. The people of Bangladesh are undergoing one of the world's largest vaccination drive. With the recent launch and introduction of the COVID-19 vaccines, many of us are curious about the general opinion or view of the vaccine. While the vaccine has ignited new hope in the battle against COVID-19, it has also sparked militant anti-vaccine campaigns, so the need to analyze public opinion on the COVID-19 vaccine has emerged. Methods: Traditional machine learning methods were used to obtain a benchmark result for the experiment. The recurrent neural network (RNN) algorithm was used next. Several different types of recurrent neural networks were used, including simple RNNs, Gated Recurrent Units (GRUs), and LSTMs. Finally, to achieve a more optimal result, small BERT models (Bidirectional Encoder Representations from Transformers) were used. Results: Upon study and testing on several models and methods, it can be seen that BERT model was the most accurate of the bunch, which was 84%. On the other hand, Naive Bayes was able to obtain an accuracy of 81%. Naive Bayes and BERT produced similar results in F1- Score, but the performance of Naive Bayes can improve as the dataset size grows. Conclusion: Knowing about public opinions on the COVID-19 vaccine is critical, and action must be taken to ensure that everybody understands the value of vaccination and that everybody receives the COVID-19 vaccine. Vaccination may help to develop immunity, which lowers the likelihood of contracting the disease and its consequences.

The dangerously contagious virus named "COVID-19" has struck the world strong and has locked down billions of people in their homes to stop the further spread. All the researchers and scientists in various fields are continually developing a vaccine and prevention methods to aid the world from this challenging situation. However, a reliable prediction of the epidemic may help control this contiguous disease until the cure is available. The machine learning techniques are one of the frontiers in predicting this outbreak's future trend and behavior. Our research is focused on finding a suitable machine learning algorithm that can predict the COVID-19 daily new cases with higher accuracy. This research has used the adaptive neuro-fuzzy inference system (ANFIS) and the long short-term memory (LSTM) to foresee the newly infected cases in Bangladesh. We have compared both the experiments' results, and it can be forenamed that LSTM has shown more satisfactory results. Upon study and testing on several models, we have shown that LSTM works better on a scenario-based model for Bangladesh with mean absolute percentage error (MAPE)-4.51, root-mean-square error (RMSE)-6.55, and correlation coefficient-0.75. This study is expected to shed light on COVID-19 prediction models for researchers working with machine learning techniques and avoid proven failures, especially for small imprecise datasets.

Robot automation technology is increasing day by day to adapt to the world's demand for controlling disasters, rescuing operations, and minimizing human risk. To perform these tasks a user-friendly robot is required which can be controlled precisely from a remote area. The described project focuses on the design and implementation of a remotely controlled bomb disposal robot with locally available hardware resources to minimize cost. The implemented robot features a robotic arm with 4 degrees of freedom which can pick and carry any sophisticated object (up to 5kg) very precisely and smoothly. The body and the arm of the robot were custom designed and built with Aluminum alloy. To maximize the capability, the robot was equipped with Gas, fire, and obstacle detection sensors. In the chassis, it can collect and carry dismantled bomb parts (up to 10kg) as evidence. The mechanical design of the robot was validated with solid works software. Proteus was used for PCB and schematic design. In the control system Raspberry pi was used as a processor and for sensing and communication Arduino mega was introduced. The robot can be controlled via the internet. An additional python script has been used to control the robot through a web server.

Objectives: The dangerously contagious virus named \newline SARS-CoV-2 has hit the world hard that has locked downed billion people in their homes for stopping further spread. All the researchers and scientists in various fields are working around the clock to come up with a vaccine and prevention methods to save the world from this invisible pathogen. However, reliable prediction of the epidemic may help contain the contagion until cure becomes available. The machine learning techniques is one of the frontier in predicting the future trend and behavior of this outbreak. Our research is focused on finding a suitable machine learning model that can predict on small dataset with higher accuracy.Methods: In this research, we have used the Adaptive Neuro-Fuzzy Inference System (ANFIS) and the long short-term memory[LSTM] to foresee the newly infected cases in Bangladesh. We have compared both the results of the experiments and it can be forenamed that LSTM has shown more satisfactory results.Results: Upon study and testing on several models, we have showed that LSTM works better on scenario based model for Bangladesh with MAPE 4.51, RMSE 6.55 and Correlation Coefficient 0.75. Conclusion: This study is expected to shade light on Covid-19 prediction models for researchers working with machine learning techniques and help avoid proven failures specially for small imprecise dataset.

Introduction: Around the world, scientists are racing hard to understand how Covid-19 epidemic is spreading and growing, thus trying to find ways to prevent it before medications come to pass. Many different models have been proposed so far correlating different factors. Some of them are too localized to indicate a general trend of the pandemic while some others have established transient correlations only.Methods: Hence, in this study, a 4P model has been proposed based on four probabilities (4P) which has been found to be true for all affected countries taking Bangladesh as a case. Efficiency scores have been estimated from survey analysis not only for governing authorities on managing the situation (P(G)) but also for the compliance of the citizens ((P(P)). Since the immune responses of all the people are not uniform to a specific pathogen, the probability of a person getting infected ((P(I)) after being exposed has also been estimated. And the vital one is the probability of Test Positivity ((P(T)) which is a strong indicator of how effectively the infected people are diagnosed and isolated from the rest of the group that affects the rate of growth.Results and Conclusion: All the four parameters have been fitted in a non-linear exponential model that partly updates itself periodically with everyday facts. Along with the model, all the four probabilistic parameters are engaged to train a recurrent neural network using Long-Short Term Memory neural network and the followed trial confirmed a ruling functionality of the 4Ps.

Introduction: Around the world, scientists are racing hard to understand how Covid-19 epidemic is spreading and growing, thus trying to find ways to prevent it before medications come to pass. Many different models have been proposed so far correlating different factors. Some of them are too localized to indicate a general trend of the pandemic while some others have established transient correlations only.Methods: Hence, in this study, a 4P model has been proposed based on four probabilities (4P) which has been found to be true for all affected countries taking Bangladesh as a case. Efficiency scores have been estimated from survey analysis not only for governing authorities on managing the situation (P(G)) but also for the compliance of the citizens ((P(P)). Since the immune responses of all the people are not uniform to a specific pathogen, the probability of a person getting infected ((P(I)) after being exposed has also been estimated. And the vital one is the probability of Test Positivity ((P(T)) which is a strong indicator of how effectively the infected people are diagnosed and isolated from the rest of the group that affects the rate of growth.Results and Conclusion: All the four parameters have been fitted in a non-linear exponential model that partly updates itself periodically with everyday facts. Along with the model, all the four probabilistic parameters are engaged to train a recurrent neural network using Long-Short Term Memory neural network and the followed trial confirmed a ruling functionality of the 4Ps.

Around the world, scientists are racing hard to understand how the COVID-19 epidemic is spreading and growing, thus trying to find ways to prevent it before medications are available. Many different models have been proposed so far correlating different factors. Some of them are too localized to indicate a general trend of the pandemic while some others have established transient correlations only. Hence, in this study, taking Bangladesh as a case, a 4P model has been proposed based on four probabilities (4P) which have been found to be true for all affected countries. Efficiency scores have been estimated from survey analysis not only for governing authorities on managing the situation (P(G)) but also for the compliance of the citizens ((P(P)). Since immune responses to a specific pathogen can vary from person to person, the probability of a person getting infected ((P(I)) after being exposed has also been estimated. And the vital one is the probability of test positivity ((P(T)) which is a strong indicator of how effectively the infected people are diagnosed and isolated from the rest of the group that affects the rate of growth. All the four parameters have been fitted in a non-linear exponential model that partly updates itself periodically with everyday facts. Along with the model, all the four probabilistic parameters are engaged to train a recurrent neural network using long short-term memory neural network and the followed trial confirmed a ruling functionality of the 4Ps.