Karpathy et al. 2014. Frizzi et al. 2016 trained CNNs to detect fire and smoke in still images. Convolutional neural networks (CNNs) have emerged as the state-of-the-art image classification algorithm due to its efficient architecture that takes advantage of the stationarity and locality of patterns found in images and videos. Unlike other machine learning methods, Convolutional Neural Networks do not rely on engineered features, but rather extract the extract the featured most relevant for classification automatically based on a labelled training set.
Convolutional neural networks (CNNs) have emerged as the state-of-the-art image classification algorithm due to its efficient architecture that takes advantage of the stationarity and locality of patterns found in images and videos. 
recognized that the existing literature was primarily rule-based models and hand-engineered features and based on the CNNs previous success with image classification, a CNN was trained to predict fire and smoke from still images. Faster R-CNN is a specific form of CNN that includes a region proposal network which hypothesizes object locations via bounding boxes more efficiently than its predecessors, R-CNN and Fast R-CNN \citealt{Ren_2017}
This project aims to create a method for detecting and recording plumes of pollution in NYC using images gathered from the Urban Observatory at New York University’s Center for Urban Science and Progress (CUSP-UO). The CUSP-UO studies the complex interactions between the physical, natural, and human components of the city as a coherent, definable system with the goal of enhancing public well-being, city operations, and future urban plans. CUSP-UO continuously images the Manhattan skyline at 0.1 Hz, and daytime images can be used to detect and characterize plumes from buildings in the scene \citealt{swurtelec2015}. The project also aims to identify various statistics such as the origin, count and frequency of the plumes. This will be performed by constructing a training dataset which will be used to train Faster R-CNN for plume location. 3D photogrammetry will then be used to identify the source building using a 3D model of the city.