The rapid growth of Internet of Things (IoT) is expected to add billions of IoT devices connected to the Internet. These devices represent a vast attack surface for cyberattacks. For example, these IoT devices can be infected with botnets to enable Distributed Denial of Service (DDoS) attacks. Signature-based intrusion detection systems are traditional countermeasures for such attacks. However, these methods rely on human experts and are time-consuming in terms of updates and may not exhaust all attack types especially zero-day attacks. Deep learning has shown some promise in intrusion detection. This paper shows that it is possible to use generative deep learning methods like Adversarial Autoencoders (AAE) and Bidirectional Generative Adversarial Networks (BiGAN) to detect intruders based on an analysis of the network data. The recently posted full IoT-23 dataset based on Somfy door lock, Philips Hue and Amazon Echo devices was used to train generative deep learning models to detect a variety of attacks like DDoS, and various botnets like Mirai, Okiruk and Torii. Over 1.8 million network flows were used to train the various models. The resulting generative models outperform traditional machine learning techniques like Random Forests. Both AAE and BiGAN-based models were able to achieve an F1-Score of 0.99. A BiGAN to detect unknown attacks was also trained to detect novel zero-day attacks with an F1-Score from 0.85 to 1.

Camera traps deployed in remote locations provide an effective method for ecologists to monitor and study wildlife in a non-invasive way. However, current camera traps suffer from two problems. First, the images are manually classified and counted, which is expensive. Second, due to manual coding, the results are often stale by the time they get to the ecologists. Using the Internet of Things (IoT) combined with deep learning represents a good solution for both these problems, as the images can be classified automatically, and the results immediately made available to ecologists. This paper proposes an IoT architecture that uses deep learning on edge devices to convey animal classification results to a mobile app using the LoRaWAN low-power, wide-area network. The primary goal of the proposed approach is to reduce the cost of the wildlife monitoring process for ecologists, and to provide real-time animal sightings data from the camera traps in the field. Camera trap image data consisting of 66,400 images were used to train the InceptionV3, MobileNetV2, ResNet18, EfficientNetB1, DenseNet121, and Xception neural network models. While performance of the trained models was statistically different (Kruskal–Wallis: Accuracy H(5) = 22.34, p < 0.05; F1-score H(5) = 13.82, p = 0.0168), there was only a 3% difference in the F1-score between the worst (MobileNet V2) and the best model (Xception). Moreover, the models made similar errors (Adjusted Rand Index (ARI) > 0.88 and Adjusted Mutual Information (AMU) > 0.82). Subsequently, the best model, Xception (Accuracy = 96.1%; F1-score = 0.87; F1-Score = 0.97 with oversampling), was optimized and deployed on the Raspberry Pi, Google Coral, and Nvidia Jetson edge devices using both TenorFlow Lite and TensorRT frameworks. Optimizing the models to run on edge devices reduced the average macro F1-Score to 0.7, and adversely affected the minority classes, reducing their F1-score to as low as 0.18. Upon stress testing, by processing 1000 images consecutively, Jetson Nano, running a TensorRT model, outperformed others with a latency of 0.276 s/image (s.d. = 0.002) while consuming an average current of 1665.21 mA. Raspberry Pi consumed the least average current (838.99 mA) with a ten times worse latency of 2.83 s/image (s.d. = 0.036). Nano was the only reasonable option as an edge device because it could capture most animals whose maximum speeds were below 80 km/h, including goats, lions, ostriches, etc. While the proposed architecture is viable, unbalanced data remain a challenge and the results can potentially be improved by using object detection to reduce imbalances and by exploring semi-supervised learning.

In recent times, transformer-based deep learning models have risen in prominence in the field of machine learning for a variety of tasks such as computer vision and text generation. Given this increased interest, a historical outlook at the development and rapid progression of transformer-based models becomes imperative in order to gain an understanding of the rise of this key architecture. This paper presents a survey of key works related to the early development and implementation of transformer models in various domains such as generative deep learning and as backbones of large language models. Previous works are classified based on their historical approaches, followed by key works in the domain of text-based applications, image-based applications, and miscellaneous applications. A quantitative and qualitative analysis of the various approaches is presented. Additionally, recent directions of transformer-related research such as those in the biomedical and timeseries domains are discussed. Finally, future research opportunities, especially regarding the multi-modality and optimization of the transformer training process, are identified.

Given recent advances in deep learning, semi-supervised techniques have seen a rise in interest. Generative adversarial networks (GANs) represent one recent approach to semi-supervised learning (SSL). This paper presents a survey method using GANs for SSL. Previous work in applying GANs to SSL are classified into pseudo-labeling/classification, encoder-based, TripleGAN-based, two GAN, manifold regularization, and stacked discriminator approaches. A quantitative and qualitative analysis of the various approaches is presented. The R3-CGAN architecture is identified as the GAN architecture with state-of-the-art results. Given the recent success of non-GAN-based approaches for SSL, future research opportunities involving the adaptation of elements of SSL into GAN-based implementations are also identified.

Maintaining biodiversity is a key component of the United Nations (UN) “Life on Land” sustainability goal. Remote camera traps monitoring animals' movements support research in biodiversity. However, images from these camera traps are currently labeled manually resulting in high processing costs and long delays. This paper proposes an IoT -based system that leverages deep learning and edge computing to automatically label camera trap images and transmit this information to scientists in a timely manner. Inception-V3, MobileNet-V2, ResNet-18, and DenseNet-121 were trained on data consisting of 33,984 images taken during day and night with 6 animal classes. Inception- V3 yielded the highest macro average F1-score of 0.93 and an accuracy of 94%. An IoT-based system was developed that directly captures images from a commercial camera trap, does the inference on the edge using a Raspberry Pi (RPi), and sends the classification results back to a cloud database system. A mobile App is used to monitor the camera images classified on camera traps in real-time. The RPi could easily sustain a rate of processing 1 image every 2 seconds with an average latency of 1.8 second/image. After capture and pre-processing, each inference took an average of 0.2 Millisecond/image on a RPi Model 4B.