An official website of the United States government
Abstract BackgroundThe research gap addressed in this study is the applicability of deep neural network (NN) models on wearable sensor data to recognize different activities performed by patients with Parkinson’s Disease (PwPD) and the generalizability of these models to PwPD using labeled healthy data. MethodsThe experiments were carried out utilizing three datasets containing wearable motion sensor readings on common activities of daily living. The collected readings were from two accelerometer sensors. PAMAP2 and MHEALTH are publicly available datasets collected from 10 and 9 healthy, young subjects, respectively. A private dataset of a similar nature collected from 14 PwPD patients was utilized as well. Deep NN models were implemented with varying levels of complexity to investigate the impact of data augmentation, manual axis reorientation, model complexity, and domain adaptation on activity recognition performance. ResultsA moderately complex model trained on the augmented PAMAP2 dataset and adapted to the Parkinson domain using domain adaptation achieved the best activity recognition performance with an accuracy of 73.02%, which was significantly higher than the accuracy of 63% reported in previous studies. The model’s F1 score of 49.79% significantly improved compared to the best cross-testing of 33.66% F1 score with only data augmentation and 2.88% F1 score without data augmentation or domain adaptation. ConclusionThese findings suggest that deep NN models originating on healthy data have the potential to recognize activities performed by PwPD accurately and that data augmentation and domain adaptation can improve the generalizability of models in the healthy-to-PwPD transfer scenario. The simple/moderately complex architectures tested in this study could generalize better to the PwPD domain when trained on a healthy dataset compared to the most complex architectures used. The findings of this study could contribute to the development of accurate wearable-based activity monitoring solutions for PwPD, improving clinical decision-making and patient outcomes based on patient activity levels.
more »
« less
Persistent monitoring of insect-pests on sticky traps through hierarchical transfer learning and slicing-aided hyper inference
IntroductionEffective monitoring of insect-pests is vital for safeguarding agricultural yields and ensuring food security. Recent advances in computer vision and machine learning have opened up significant possibilities of automated persistent monitoring of insect-pests through reliable detection and counting of insects in setups such as yellow sticky traps. However, this task is fraught with complexities, encompassing challenges such as, laborious dataset annotation, recognizing small insect-pests in low-resolution or distant images, and the intricate variations across insect-pests life stages and species classes. MethodsTo tackle these obstacles, this work investigates combining two solutions, Hierarchical Transfer Learning (HTL) and Slicing-Aided Hyper Inference (SAHI), along with applying a detection model. HTL pioneers a multi-step knowledge transfer paradigm, harnessing intermediary in-domain datasets to facilitate model adaptation. Moreover, slicing-aided hyper inference subdivides images into overlapping patches, conducting independent object detection on each patch before merging outcomes for precise, comprehensive results. ResultsThe outcomes underscore the substantial improvement achievable in detection results by integrating a diverse and expansive in-domain dataset within the HTL method, complemented by the utilization of SAHI. DiscussionWe also present a hardware and software infrastructure for deploying such models for real-life applications. Our results can assist researchers and practitioners looking for solutions for insect-pest detection and quantification on yellow sticky traps.
more »
« less
- Award ID(s):
- 1954556
- PAR ID:
- 10579967
- Publisher / Repository:
- Frontiers in Plant Science
- Date Published:
- Journal Name:
- Frontiers in Plant Science
- Volume:
- 15
- ISSN:
- 1664-462X
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Abstract Insect pests cause significant damage to food production, so early detection and efficient mitigation strategies are crucial. There is a continual shift toward machine learning (ML)‐based approaches for automating agricultural pest detection. Although supervised learning has achieved remarkable progress in this regard, it is impeded by the need for significant expert involvement in labeling the data used for model training. This makes real‐world applications tedious and oftentimes infeasible. Recently, self‐supervised learning (SSL) approaches have provided a viable alternative to training ML models with minimal annotations. Here, we present an SSL approach to classify 22 insect pests. The framework was assessed on raw and segmented field‐captured images using three different SSL methods, Nearest Neighbor Contrastive Learning of Visual Representations (NNCLR), Bootstrap Your Own Latent, and Barlow Twins. SSL pre‐training was done on ResNet‐18 and ResNet‐50 models using all three SSL methods on the original RGB images and foreground segmented images. The performance of SSL pre‐training methods was evaluated using linear probing of SSL representations and end‐to‐end fine‐tuning approaches. The SSL‐pre‐trained convolutional neural network models were able to perform annotation‐efficient classification. NNCLR was the best performing SSL method for both linear and full model fine‐tuning. With just 5% annotated images, transfer learning with ImageNet initialization obtained 74% accuracy, whereas NNCLR achieved an improved classification accuracy of 79% for end‐to‐end fine‐tuning. Models created using SSL pre‐training consistently performed better, especially under very low annotation, and were robust to object class imbalances. These approaches help overcome annotation bottlenecks and are resource efficient.more » « less
-
Abstract Long‐term experiments are critical for understanding ecological processes, but their management comes with unique challenges. As time passes, projects may encounter unavoidable changes due to external factors, like availability of materials, affecting aspects of their research methodology. At the Kellogg Biological Station Long‐Term Ecological Research Site, one of the many National Science Foundation‐funded long‐term research stations, a three‐decade project recently experienced a supply‐chain‐induced change in insect sampling methodology in their lady beetle observation study. Since 1989, lady beetles (Coleoptera: Coccinellidae) have been sampled weekly over the growing season using yellow sticky cards. In 2021, the original sticky traps were discontinued by the manufacturer and replaced with a similar, but not identical trap. We conducted a 3‐year study while the new traps were phased in to examine how the trap change would impact the observed biodiversity patterns at the site. We examined community metrics and individual taxa captures to examine within‐year and between‐year differences in performance between the card types. Overall, we noted several small but statistically detectable differences in capture patterns between the two trap types. After accounting for other sources of variation, we observed a difference in Shannon diversity of insects captured on the two card types, but not richness or abundance, for the overall insect community. Yet, these differences were dwarfed by the magnitude of difference observed between years within card types. For individual taxa, similar patterns held: between trap differences could be detected statistically, but the number of differences in capture rate between trap types was less than the number of differences observed for the same trap, between years. Thus, we conclude that while subtle changes in methodology could impact data produced in long‐term experiments; in this case, the magnitude of this change is smaller than other factors such as time and plant treatment. However, if sustained changes in the capture rates of focal taxa are observed, future data users may use our observations to specifically quantify and correct for these shifting patterns related to the protocol change.more » « less
-
Monitoring and analysis of wildlife are key to conservation planning and conflict management. The widespread use of camera traps coupled with AI-based analysis tools serves as an excellent example of successful and non-invasive use of technology for design, planning, and evaluation of conservation policies. As opposed to the typical use of camera traps that capture still images or short videos, in this project, we propose to analyze longer term videos monitoring a large flock of birds. This project, which is part of the NSF-TIH Indo-US joint R&D partnership, focuses on solving challenges associated with the analysis of long-term videos captured at feeding grounds and nesting sites, among other such locations that host large flocks of migratory birds. We foresee that the objectives of this project would lead to datasets and benchmarking tools as well as novel algorithms that would be instrumental in developing automated video analysis tools that could in turn help understand individual and social behavior of birds. The first of the key outcomes of this research will include the curation of challenging, real-world datasets for benchmarking various image and video analytics algorithms for tasks such as counting, detection, segmentation, and tracking. Our recent efforts towards this outcome is a curated dataset of 812 high-resolution, point-annotated, images (4K - 32MP) of a flock of Demoiselle cranes (Anthropoides virgo) taken from their feeding site at Khichan, Rajasthan, India. The average number of birds in each image is about 207, with a maximum count of 1500. The benchmark experiments show that state-of-the-art vision techniques struggle with tasks such as segmentation, detection, localization, and density estimation for the proposed dataset. Over the execution of this open science research, we will be scaling this dataset for segmentation and tracking in videos, as well as developing novel techniques for video analytics for wildlife monitoring.more » « less
-
Abstract ObjectivesPregnancy failure represents a major fitness cost for any mammal, particularly those with slow life histories such as primates. Here, we quantified the risk of fetal loss in wild hybrid baboons, including genetic, ecological, and demographic sources of variance. We were particularly interested in testing the hypothesis that hybridization increases fetal loss rates. Such an effect would help explain how baboons may maintain genetic and phenotypic integrity despite interspecific gene flow. Materials and MethodsWe analyzed outcomes for 1020 pregnancies observed over 46 years in a natural yellow baboon‐anubis baboon hybrid zone. Fetal losses and live births were scored based on records of female reproductive state and the appearance of live neonates. We modeled the probability of fetal loss as a function of a female's genetic ancestry (the proportion of her genome estimated to be descended from anubis [vs. yellow] ancestors), age, number of previous fetal losses, dominance rank, group size, climate, and habitat quality using binomial mixed effects models. ResultsFemale genetic ancestry did not predict fetal loss. Instead, the risk of fetal loss is elevated for very young and very old females. Fetal loss is most robustly predicted by ecological factors, including poor habitat quality prior to a home range shift and extreme heat during pregnancy. DiscussionOur results suggest that gene flow between yellow and anubis baboons is not impeded by an increased risk of fetal loss for hybrid females. Instead, ecological conditions and female age are key determinants of this component of female reproductive success.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.3389/fpls.2024.1484587
