An official website of the United States government
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
The Importance of the Instantaneous Phase for Face Detection using Simple Convolutional Neural Networks
Large scale training of Deep Learning methods requires significant computational resources. The use of transfer learning methods tends to speed up learning while producing complex networks that are very hard to interpret. This paper investigates the use of a low-complexity image processing system to investigate the advantages of using AM-FM representations versus raw images for face detection. Thus, instead of raw images, we consider the advantages of using AM, FM, or AM-FM representations derived from a low-complexity filterbank and processed through a reduced LeNet-5. The results showed that there are significant advantages associated with the use of FM representations. FM images enabled very fast training over a few epochs while neither IA nor raw images produced any meaningful training for such low-complexity network. Furthermore, the use of FM images was 7x to 11x faster to train per epoch while using 123x less parameters than a reduced-complexity MobileNetV2, at comparable performance (AUC of 0.79 vs 0.80).
more »
« less
- PAR ID:
- 10184950
- Date Published:
- Journal Name:
- 2020 IEEE Southwest Symposium on Image Analysis and Interpretation (SSIAI)
- Page Range / eLocation ID:
- 1 to 4
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Additive manufacturing (AM) methods have become mainstream in many industry sectors, especially aeronautics and space structures, where production volume for components is low and designs are highly customized. The frequency of launching space missions is increasing around the world. Some of these missions are sending landers and rovers to moon, mars, and other planets. Such space structures require numerous parts that are unique in design or are produced in just one or a very small production run. Such parts produced for high stake and very expensive missions require complete confidence in the quality of each part. Characterization of parts manufactured by AM is a significant challenge for many existing methods due to the geometric complexity, feature size in the structure, and size of the part. This paper discusses various challenges in applying current characterization methods to the AM sector. Machine learning (ML) methods are considered promising in materials and manufacturing fields. However, generating the training dataset by creating a large number of parts is expensive and impractical. New methods are required to train the ML algorithms on small datasets, especially for parts of unique geometry that are produced in limited production run such as space structures.more » « less
-
Abstract Representations of the world environment play a crucial role in artificial intelligence. It is often inefficient to conduct reasoning and inference directly in the space of raw sensory representations, such as pixel values of images. Representation learning allows us to automatically discover suitable representations from raw sensory data. For example, given raw sensory data, a deep neural network learns nonlinear representations at its hidden layers, which are subsequently used for classification (or regression) at its output layer. This happens implicitly during training through minimizing a supervised or unsupervised loss. In this letter, we study the dynamics of such implicit nonlinear representation learning. We identify a pair of a new assumption and a novel condition, called the on-model structure assumption and the data architecture alignment condition. Under the on-model structure assumption, the data architecture alignment condition is shown to be sufficient for the global convergence and necessary for global optimality. Moreover, our theory explains how and when increasing network size does and does not improve the training behaviors in the practical regime. Our results provide practical guidance for designing a model structure; for example, the on-model structure assumption can be used as a justification for using a particular model structure instead of others. As an application, we then derive a new training framework, which satisfies the data architecture alignment condition without assuming it by automatically modifying any given training algorithm dependent on data and architecture. Given a standard training algorithm, the framework running its modified version is empirically shown to maintain competitive (practical) test performances while providing global convergence guarantees for deep residual neural networks with convolutions, skip connections, and batch normalization with standard benchmark data sets, including MNIST, CIFAR-10, CIFAR-100, Semeion, KMNIST, and SVHN.more » « less
-
We introduce the problem of detecting a group of students from classroom videos. The problem requires the detection of students from different angles and the separation of the group from other groups in long videos (one to one and a half hours). We use multiple image representations to solve the problem. We use FM components to separate each group from background groups, AM-FM components for detecting the back-of-the-head, and YOLO for face detection. We use classroom videos from four different groups to validate our approach. Our use of multiple representations is shown to be significantly more accurate than the use of YOLO alone.more » « less
-
In this research paper, we compare the alignment between advanced manufacturing (AM) competencies in Florida’s Career and Technical Education (CTE) AM Curriculum Framework and the U.S. Department of Labor’s Advanced Manufacturing Competency Model. AM educators are guided by state department of education documents that specify program content, while employers track the knowledge, skills, and dispositions that AM technicians require to successfully function in the workplace. The Curriculum Framework, created with input from educators and industry, shape AM curricula and course syllabi because they specify the learning outcomes that AM graduates upon completion of two-year AM degree programs. The Department of Labor’s Advanced Manufacturing Competency Model, crafted by federal policymakers and industry representatives, includes personal, academic, industry-specific, and managerial competencies needed by successful AM technicians; the Model is intended to influence technicians’ hiring, training, and evaluation. Although these documents were created by different sets of stakeholders, they “bookend” AM technicians’ school-to-career pathways. To determine the extent to which the 2019-2020 Florida AM Curriculum Framework aligns to the Department of Labor’s Advanced Manufacturing Competency Model, we used text mining to extract and compare the key competencies found in both documents. Through this approach, we compared these documents and identified: 1) frequently addressed topics; 2) verbs that guided the complexity (i.e., Bloom’s Revised Taxonomy of Learning Objectives cognitive level) of the course learning task versus workplace competency; and 3) overall match between the documents. Our results suggest that the documents overlap very little, with significant misalignments in higher-level Bloom’s verbs. We present implications for educational institutions, AM policy makers, and industry; suggest a revision cycle and process; and propose an ongoing assessment model to improve the congruence between what employers want and what is taught in two-year AM degree programs.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- https://doi.org/10.1109/SSIAI49293.2020.9094589
