The recognition of behavioral relationships related to aggression in pigs is an important task that is visually observed in the livestock industry. However, this task is laborious and susceptible to errors, which can be reduced through automation by visually classifying videos captured in a controlled environment. Video classification can be automated using techniques from computer vision and artificial intelligence, employing neural network methods. The primary techniques utilized in this study are variants of transformers: STAM, TimeSformer, and ViViT, as well as techniques using convolutions, such as ResNet3D2, ResNet(2+1)D, and CnnLstm. These techniques were compared to analyze their individual contributions. The performance was evaluated using metrics such as accuracy, precision, and recall. The TimeSformer technique demonstrated the most promising results in video classification, achieving a median accuracy of 0.729.

Smart indoor tourist attractions, such as smart museums and aquariums, require a significant investment in indoor localization devices. The use of Global Positioning Systems on smartphones is unsuitable for scenarios where dense materials such as concrete and metal blocks weaken GPS signals, which is most often the case in indoor tourist attractions. With the help of deep learning, indoor localization can be done region by region using smartphone images. This approach requires no investment in infrastructure and reduces the cost and time needed to turn museums and aquariums into smart museums or smart aquariums. In this paper, we propose using deep learning algorithms to classify locations based on smartphone camera images for indoor tourist attractions. We evaluate our proposal in a real-world scenario in Brazil. We extensively collect images from ten different smartphones to classify biome-themed fish tanks in the Pantanal Biopark, creating a new dataset of 3654 images. We tested seven state-of-the-art neural networks, three of them based on transformers. On average, we achieved a precision of about 90% and a recall and f-score of about 89%. The results show that the proposal is suitable for most indoor tourist attractions.

Aedes aegypti is still one of the main concerns when it comes to disease vectors. Among the many ways to deal with it, there are important protocols that make use of egg numbers in ovitraps to calculate indices, such as LIRAa and Breteau Index, which can provide information on predictable outbursts and epidemics. Also, there are many research lines that require egg numbers, specially when mass production of mosquitoes is needed. Egg counting is a laborious and error-prone task that can be automated via computer vision-based techniques, specially deep learning-based counting with object detection. In this work, we propose a new dataset comprising field and laboratory eggs, along with test results of three neural networks applied to the task: Faster R-CNN, Side-Aware Boundary Localization and FoveaBox. With FoveaBox, we achieve a median mean absolute error of 6.854. Finally, we also discuss the main difficulties and possibilities for future research.

Forest inventory is an important activity for planning and decision-making in forest management. It is usually carried out in the field using different sampling methods and processes, which are usually limited by its high costs and by the scarcity of manpower. In this work, we evaluate deep learning methods in qualitative forest inventory of Eucalyptus plantations using unmanned aerial vehicles, multispectral sensors and deep learning. For the evaluation, we present a dataset collected in two study areas located in different municipalities in the State of Mato Grosso do Sul, including field measurements collected by occasion of the qualitative forest inventory at 4 months (QFI 4m) and aerophotogrammetric coverage of 36 plots represented by 124 sampling units. State-of-the-art neural networks were then used to predict four variables, collected through traditional QFI 4m and approximated by two models: PB50 and PC50, which are adaptations of the PV50 index, and the total and average biomass in the sampling unit. The results show that the transformer-based architecture MaxViT presented the lowest errors in predicting all the variables. For example, for the PB50 variable, it achieved an RMSE of 7.5 (± 1.85) and a MAPE of 0.33 (± 0.23).