Paper Title
Identification of Objects for Visually Impaired Persons Using CNN

As object recognition technology has developed recently, various technologies have object detection technique for blind people in real time to detect objects on any device running this model. We use Convolutional neural network along with single shot multi-box detector algorithm to develop the proposed model. This model is composed of multiple layers to classify the given objects into any of the defined classes. Due to recent advancement in deep learning with image processing, enable us to develop this model. Our model takes color images as input and train the mode until the error rate is less. We have tested our model by supplying some sample images. To increase the computational performance of the model we have used single shot multi-box detector algorithm along with the help of architecture of faster region Convolutional Neural Network. We also have calculated the accuracy for detecting the objects and found that the accuracy is acceptable compare to other existing scheme. For that, we have considered different parameter like mean average precision (mAP) and frames per second (FPS). The single shot multi-box detector algorithm uses standard VOC and COCO datasets. This model integrates the audio device, which will be helpful for the blind people. Keywords - Object detection, Voice output, Python OpenCV, CNN