A Review: Image Classification and Object Detection with Deep Learning

Deep learning has been developed as an efficient machine learning approach that incorporates several layers of features or data interpretation and delivers state-of-the-art results. The application of deep learning has illustrated amazing performance in various fields of implementation, especially in the field of image classification, segmentation, and object detection. Recent developments in deep learning methods are promoting efficiency in fine-grained image classification, which intends to differentiate between sub-categories. In this paper, we include a thorough analysis of the different deep architectures and frameworks illustrating the model specifications. CNN or convolutional neural network has been the fundamental way to go for object detection computer vision and much more. But with the development of data into a more complex form, classical CNN is not capable to provide up to the mark results. Hence, this review paper is made with the aim to bring some prominent models and techniques back into the light and provide their results on different popular datasets. There have been key discoveries made which are discussed throughout the paper.