Find research datasets worth reusing

Fig. 9 in An optimised YOLOv4 deep learning model for efficient malarial cell detection in thin blood smear images

Fig. 9 Overall framework of proposed automated malaria diagnosis and species identification. CNN, Convolutional neural network; RBC, red blood cell; YOLO, You Only Look Once (model)

Fig. 8 in An optimised YOLOv4 deep learning model for efficient malarial cell detection in thin blood smear images

Fig. 8 Examples of false positive predictions by the YOLOv4-RC3_4 model. YOLO, You Only Look Once (model)

Fig. 6 in An optimised YOLOv4 deep learning model for efficient malarial cell detection in thin blood smear images

Fig. 6 Comparison of detection performance by the original YOLOv4 model and the YOLOv4-RC3_4 model. Red arrows indicate cells not detected by the original YOLOv4 model, green arrows indicate the same cells detected by the YOLOv4-RC3_4 model. YOLO, You Only Look Once (model)

Fig. 3 in An optimised YOLOv4 deep learning model for efficient malarial cell detection in thin blood smear images

Fig. 3 Network structure of YOLOv4. CSP, cross-spatial connection; SPP, spatial pyramid pooling layer; PANet, Path Aggregation Network; CBM, Convolutional, Batch Normalisation, and Activation; CBL, Convolutional, Batch normalisation, and Leaky-ReLU; Conv, convolutional; Concat, concatenation

Fig. 4 in An optimised YOLOv4 deep learning model for efficient malarial cell detection in thin blood smear images

Fig. 4 Building blocks of the residual learning module. CBM, Convolutional, Batch normalisation and Mish (modules)

Fig. 5 in An optimised YOLOv4 deep learning model for efficient malarial cell detection in thin blood smear images

Fig. 5 Visual representation of the removal of residual blocks from C3 and C4 Res-block body. YOLO,You Only Look Once (model)

Fig. 1 in An optimised YOLOv4 deep learning model for efficient malarial cell detection in thin blood smear images

Fig. 1 Comparison of malaria diagnosis using deep learning CNN models and deep learning object detectors. CNN, Convolutional neural network

Fig. 2 in An optimised YOLOv4 deep learning model for efficient malarial cell detection in thin blood smear images

Fig. 2 Cropping of infected cells using the coordinates of predictions by the object detectors. RBC, Red blood cell; YOLO,You Only Look Once (model)

opencc-by-4.0May 2024View details →

Cloud to Thing Continuum based Sports Monitoring System using Machine Learning and Deep Learning Model

Sports monitoring and analysis have seen significant advancements with the integration of cloud computing and continuum paradigms, facilitated by machine learning and deep learning techniques. In this study, we present a novel approach for sports monitoring that seamlessly transitions from traditional cloud-based architectures to a continuum paradigm, enabling real-time analysis and insights into player performance and team dynamics. Leveraging machine learning and deep learning algorithms, our framework offers enhanced capabilities for player tracking, action recognition, and performance evaluation in various sports scenarios. This research proposes a Cloud-to-Thing Continuum based Sports Monitoring System utilizing Machine Learning (ML) and Deep Learning (DL) models. The system integrates data acquisition, preprocessing, feature extraction, cloud-based processing, continuum paradigm integration, and decision-making stages. It leverages innovative techniques such as Improved Mask R-CNN for pose estimation, hybrid metaheuristic algorithms with Generative Adversarial Network (GAN) for classification, and fuzzy decision-making Based on the integrated analysis, decisions are made regarding player performance, team strategies, and tactical adjustments. The continuum approach ensures a balance between centralized cloud processing and distributed edge processing, optimizing resource utilization and reducing latency. Through this system, real-time analysis of sports events is achieved, enabling immediate feedback for time-sensitive applications.

opencc-by-4.0May 2024View details →

Replication Data and Analyses for: J. Monsimet, S. Sjögersten, N.J. Sanders, M. Jonsson, J. Olofsson & M. Siewert, 2024. UAV data and deep learning: efficient tools to map the ecological footprint of ants mounds, Remote Sensing in Ecology and Conservation.

This dataset corresponds to the article: "Jérémy Monsimet*¹, Sofie Sjögersten², Nathan J. Sanders³, Micael Jonsson¹, Johan Olofsson¹, Matthias Siewert¹, 2024. UAV data and deep learning: efficient tools to map the ecological footprint of ants mounds, Remote Sensing in Ecology and Conservation" DOI: <a href="https://doi.org/10.1002/rse2.400" target="_blank" rel="nofollow noreferrer noopener">10.1002/rse2.400</a> 1 Department of Ecology and Environmental Science, Umeå University, Sweden 2 School of Biosciences, University of Nottingham, Loughborough, UK 3 Department of Ecology and Evolutionary Biology, University of Michigan, US The gitlab repository of this dataset is available at: <a href="https://gitlab.com/Monsimet/uav_ants_treeline/-/tree/main/">https://gitlab.com/Monsimet/uav_ants_treeline/-/tree/main/</a> In this repository, you will find the analyses and results presented in the paper. In each folder, there is a html file that can be read after downloading locally the whole folder. You can either run the .qmd file used to produce the html file or walk through the html files (see the readme.md for more information). Paper abstract: High‐resolution unoccupied aerial vehicle (UAVs) data have alleviated the mismatch between the scale of ecological processes and the scale of remotely sensed data, while machine learning and deep learning methods allow new avenues for quantification in ecology. Ant nests play key roles in ecosystem functioning, yet their distribution and effects on entire landscapes remain poorly understood, in part because they and their mounds are too small for satellite remote sensing. This research maps the distribution and impact of ant mounds in a 20 ha treeline ecotone. We evaluate the detectability from UAV imagery using a deep learning model for object detection and different combinations of RGB, thermal and multispectral sensor data. We were able to detect ant mounds in all imagery using manual detection and deep learning. However, the highest precision rates were achieved by deep learning using RGB data which has the highest spatial resolution (1.9 cm) at comparable UAV flight height. While multispectral data were outperformed for detection, it allows for novel insights into the ecology of ants and their spatial impact on vegetation productivity using the normalized difference vegetation index. Scaling up, this suggests that ant mounds quantifiably impact vegetation productivity for up to 4% of our study area and up to 8% of the Betula nana vegetation communities, the vegetation type with the highest abundance of ant mounds. Therefore, they could have an overlooked role in nutrient‐limited tundra vegetation, and on the shrubification of this habitat. Further, we show the powerful combination UAV multi‐sensor data and deep learning for efficient ecological tracking and monitoring of mound‐building ants and their spatial impact.

opencc-by-4.0Mar 2024View details →

Dataset for "Solving deep-learning density-functional theory via variational autoencoder"

The dataset contains the ground state energies, the ground state density profiles, and the external potentials of a 3D single particle system with a Gaussian-like external potential. The number of grid points for each dimension is \(N_g=18\), the linear length of the box is \(L=a_0\) with \(a_0\) the unit of length. The unit of energy is \(E_0=\frac{ \hbar^2}{(m a_0^2)}\).     The dataset is zip file of a Python npz file with the following keys: - "density" that corresponds to the ground state density profile. - "potential" is the external potential. -"energy" is the ground state energy. The number of instances is 36000.  -3D_gaussian.zip -> 3D_gaussian.npz     a dictionary with three keys -density, potential, energy-.     The dimension of both potential and density is \([N_d,N_g,N_g,N_g]\).     The shape of energy is \([N_d]\).     \(N_d=36000\) -3D_gaussian_transfer_test_1.npz     a dictionary with three keys -density, potential, energy-.     The dimension of both potential and density is \([N_d,N_g,N_g,N_g]\).     The shape of energy is \([N_d]\).     \(N_d=500\) -3D_gaussian_transfer_test_2.npz     a dictionary with three keys -density, potential, energy-.     The dimension of both potential and density is \([N_d,N_g,N_g,N_g]\).     The shape of energy is \([N_d]\).     \(N_d=500\)

opencc-by-4.0May 2024View details →

The Potential of Deep Learning Object Detection in Citizen-Driven Snail Host Monitoring to Map Putative Disease Transmission Sites

<a name="_Hlk158802145"></a>Schistosomiasis is a neglected tropical disease caused by parasitic flukes transmitted by freshwater snails. Despite increasing efforts of mass drug administration, schistosomiasis remains a public health concern and the World Health Organization recommends complementary snail control. To address the need of broad-scale and actual snail distribution data to guide snail control, we adopted a citizen science approach and recruited citizen scientists (CS) to perform weekly snail sampling in the endemic setting in Uganda. Snails were identified, sorted and counted according to genus, photographed and uploaded for expert-led validation and feedback. However, expert validation is time-consuming and introduces a delay in verified data output. Thus, artificial intelligence could provide a solution by means of automated detection and counting of multiple snails collected from the field. Trained on approximately 2500 citizen-collected images, the resulting model can simultaneously detect and count Biomphalaria and Radix snails with average precision of 98.1% and 98.8% respectively. The object detection model also agreed with the expert’s decision averagely for 98.8% of the test images and could be ran in real-time (24.6 images per second). We conclude that the automatic and instant detection can rapidly and reliably validate data submitted by CS in the field, ultimately minimizing the expert validation efforts and thereby facilitating the mapping of putative schistosomiasis transmission sites. An extension to a mobile application could equip citizen scientists in remote areas with instant learning opportunities and expert-like identification skills, overcoming the need for on-site training and extensive expert intervention.

dryad40/100

Data and code from: Learning a deep language model for microbiomes: The power of large scale unlabeled microbiome data

We use open source human gut microbiome data to learn a microbial "language" model by adapting techniques from Natural Language Processing (NLP). Our microbial "language" model is trained in a self-supervised fashion (i.e., without additional external labels) to capture the interactions among different microbial species and the common compositional patterns in microbial communities. The learned model produces contextualized taxa representations that allow a single bacteria species to be represented differently according to the specific microbial environment it appears in. The model further provides a sample representation by collectively interpreting different bacteria species in the sample and their interactions as a whole. We show that, compared to baseline representations, our sample representation consistently leads to improved performance for multiple prediction tasks including predicting Irritable Bowel Disease (IBD) and diet patterns. Coupled with a simple ensemble strategy, it produces a highly robust IBD prediction model that generalizes well to microbiome data independently collected from different populations with substantial distribution shift. We visualize the contextualized taxa representations and find that they exhibit meaningful phylum-level structure, despite never exposing the model to such a signal. Finally, we apply an interpretation method to highlight bacterial species that are particularly influential in driving our model's predictions for IBD.

opencc-zeroJun 2024View details →

opencc-by-4.0Apr 2022View details →

BIM-Speed training dataset for HVAC detection using Deep Learning

Training dataset formed by two types of elements (for now) to be detected by the model using Deep Learning: Boiler and Radiator. Of the boiler type 107 images have been collected and of the radiator type 123 images. Although there are few images to get good training, it can be used as a first approximation and to evaluate the precision obtained with few training images. These images have been distributed in 80% of the images for the training phase and 20% for the test phase. All the images have the bounding box data for each element in the xml file with the same name. Additionally, 33 other images (not previously used) have been used for the validation of the tool. More information can be found here (subsection 3.1 Detection and classification of HVAC components in images): https://www.bim-speed.eu/en/Project%20Results%20%20Documents/Deliverables/D1.2.pdf

opencc-by-4.0Jul 2024View details →

Datasets for "Deep learning-based design of synthetic orthologs of SH3 signaling domains"

Description of data for "Deep learning-based design of synthetic orthologs of SH3 signaling domains": biochemistry_data.zip --> contains the binding assay, melting temperature, and enthalpy measurements that reproduce table 1 in the main text. sequence_data.zip --> contains the sequences with relative enrichment measurements and other meta data information (e.g. latent embeddings, paralog labels, etc.). sequence_data.zip > SH3_Library_Natural.xlsx --> contains the natural alleles with normalized relative enrichment scores, paralog labels, mmd latent coordinates, and among other meta data. sequence_data.zip > SH3_Library_Design.xlsx --> contains the design alleles with normalized relative enrichment scores, mmd latent coordinates, and among other meta data. sequence_data.zip > paralog_mapping.xlsx --> contains the mapping between paralog name and COG labels. note: to access these spreadsheets for analysis, we recommend using pandas library in Python. To reproduce figures within the manuscript, please follow this github repo link: https://github.com/chemgeeklian/SH3_orthology_paper_analysis

Figure 2 in Using deep-learning for automatic identification of images of marine benthic macro-invertebrate bycatch: a proof of concept

Figure 2. – Three images of organisms obtained by cropping images of lots; from left to right: Chalinidae (Porifera), Polyclinidae (Chordata), Hormatidae (Cnidaria).

Figure 3 in Using deep-learning for automatic identification of images of marine benthic macro-invertebrate bycatch: a proof of concept

Figure 3. – Image of a batch of macro-invertebrate bycatch organisms from Kerguelen Exclusive Economical Zone (Poker 4 survey, 2017), including corals, a crinoïd, an ophiurid, a sea urchin and a brachiopoda; organisms are incomplete and have been quickly spread out over a small plate to take the picture.

Figure 6 in Using deep-learning for automatic identification of images of marine benthic macro-invertebrate bycatch: a proof of concept

Figure 6. – Example of detection and classification obtained with an image including an Ophiuroid, a piece of coral and a sea star with network 2; red squares and annotations have been provided by the computer with no human action.

Figure 5 in Using deep-learning for automatic identification of images of marine benthic macro-invertebrate bycatch: a proof of concept

Figure 5. – Example of detection and classification obtained with an image including Ascidians and a sea star with network 2; red squares and annotations have been provided by the computer with no human action.