Cost Effective Obstacle Detection Kit for Blind People Akanksha Mishra, Dev Sparsh Sangwan, Saigeeta Priyadarshini, Ankit Kumar Jaiswal, Lavish Kumar Singh Icdt 2025 3rd International Conference on Disruptive Technologies, 2025 Around 1.3 billion people worldwide suffer from various types of vision impairments, and most of them encounter accidents due to obstacles they are unaware of. Most of the proposed solutions in this regard require a continuous network connection, are difficult to produce and design, require the user to carry a smartphone, and are costly. The present research aims to provide an effective, easy-to-use, and cost-efficient solution. The kit design primarily consists of ultrasonic sensors, an Arduino Uno processor, and micro-motors. Ultrasonic sensors were placed on the legs, hands, and cap of multiple users. As soon as obstacles entered the sensing range, signals were generated and sent to the Arduino Uno controller for processing. The Arduino Uno then processed the signals and used the micro-motors to create the vibratory output. The developed obstacle detection kit offers several advantages, including low cost, ease of adaptation and maintenance, a simple design, the utilization of the human body for an improved working range, and, most importantly, the ability to detect objects not only on the ground but also suspended ones.
Multi-Layer Feature Fusion-based Deep Multi-layer Depth Separable Convolution Neural Network for Alzheimer's Disease Detection Santosh Kumar Tripathy, Divya Singh, Ankit Jaiswal 2023 International Conference on Iot Communication and Automation Technology Icicat 2023, 2023 Alzheimer's disease (AD) is a severe degenerative neurological disorder that can cause heart and respiratory dysfunction. Thus, early detection of such disease is highly required. Using MRI images, a number of deep models have been created to predict AD as artificial intelligence (AI) has advanced. However, these models suffer from limited representation in extracting fine-grained features from MRI images thereby performance is declined. The proposed model overcomes such limitation and presents a methodology for AD detection by proposing a novel multi-layer feature fusion-based deep multi-layer depth-wise separable convolution neural network (CNN). The proposed model enhances the quality of features by fusing multi-layer features. These features range from representations of low-level features to features at the object level. The fused multiscale features are used for predicting AD disease. Publicly available dataset is used to validate the model's performance. With an accuracy of 95.16 percent, the suggested model performs better than the current state of the art.
FSD: A novel forged document dataset and baseline Ankit Kumar Jaiswal, Shiksha Singh, Santosh Kumar Tripathy 2023 14th International Conference on Computing Communication and Networking Technologies Icccnt 2023, 2023 Forgery detection in multimedia is an emerging research area in the field of digital content security. The current research focuses on finding several solutions to digital image forgery detection but a small number of works have been reported for scanned document forgery detection. The reason would be the unavailability of a forged scanned document (FSD) dataset. A scanned document can be manipulated in different ways: deleting a region, copy paste a region of the document either in the same document or in another document. Detection of these manipulations becomes arduous when a post-processing operation is performed on the manipulated region. To advance research in this field, we created a large dataset for the FSD. To generate such a dataset, we collected scanned documents from the FUNSD dataset (Publicly Available). The resolution of each sample of the dataset is of size 755x1000. The samples of the dataset are forged using different post-processing operations. The dataset comprises 6656 instances of manipulated scanned documents with their ground truth masks. To build the baseline for forged scanned document detection, we evaluated state-of-the-art object segmentation algorithms on the FSD dataset. Experiments demonstrate that the FSD well represents forged scanned documents and is quite challenging.
