Research has shown that climate change creates warmer temperatures and drier conditions, leading to longer wildfire seasons and increased wildfire risks in the United States. These factors have, in turn, led to increases in the frequency, extent, and severity of wildfires in recent years. Given the danger posed by wildland fires to people, property, wildlife, and the environment, there is an urgent need to provide tools for effective wildfire management. Early detection of wildfires is essential to minimizing potentially catastrophic destruction. To that end, in this paper, we present our work on integrating multiple data sources into SmokeyNet, a deep learning model using spatiotemporal information to detect smoke from wildland fires. We present Multimodal SmokeyNet and SmokeyNet Ensemble for multimodal wildland fire smoke detection using satellite-based fire detections, weather sensor measurements, and optical camera images. An analysis is provided to compare these multimodal approaches to the baseline SmokeyNet in terms of accuracy metrics, as well as time-to-detect, which is important for the early detection of wildfires. Our results show that incorporating weather data in SmokeyNet improves performance numerically in terms of both F1 and time-to-detect over the baseline with a single data source. With a time-to-detect of only a few minutes, SmokeyNet can be used for automated early notification of wildfires, providing a useful tool in the fight against destructive wildfires.

Pancreatic cancer is one of the leading causes of deaths. At later stages, it is impossible to cure. However, if the early symptoms of cancer are identified using diagnostic methods, there is a possibility of treatment and gradual cure. In order to aid in the diagnosis of cancer of the pancreas, the application of Artificial Intelligence was viewed as a solution. Initially, various machine learning algorithms like Support Vector Machines, Twin Support Vector Machines and RadSTM-ER were used, yielding encouraging results. However, the arduous process of feature extraction and manual feature selection resulted in researchers turning towards deep learning methods. This research work uses ultrasound images of the pancreas obtained from an open sourced database by the gastroenterology unit of National University of Colombia. The implementation was carried out using transfer learning on Densely Connected Convolutional Neural Networks (DenseNets). DenseNet201 using pretrained weights of ImageNet was fine-tuned on 26469 pancreatic ultrasound images consisting of three categories: cancer, healthy and pancreatitis. Additionally, cross validation on 2940 images yielded an accuracy of 99.87 %. Apart from this, 9804 images of the same distribution were used to test the performance. The DenseNet201 model was able to yield an accuracy of 99.877 %, a sensitivity of 0.9987, a specificity of 0.99931 and a positive prediction value of 0.99877, with only 12 out of 9804 misclassified examples. The ability of deep transfer learning on DenseNet201 to generalize exceedingly well encourages its beta testing on real time data and subsequent deployment.