More Like this

1. Evolutionary Programming Based Deep Feature and Model Selection for Visual Data Classification

   https://doi.org/10.1109/MIPR49039.2020.00020  

   Tian, Haiman; Chen, Shu-Ching; Shyu, Mei-Ling (August 2020, 2020 IEEE Conference on Multimedia Information Processing and Retrieval (MIPR))

   null (Ed.)

   Abstract: Deep Learning (DL) has made significant changes to a large number of research areas in recent decades. For example, several astonishing Convolutional Neural Network (CNN) models have been built by researchers to fulfill image classification needs using large-scale visual datasets successfully. Transfer Learning (TL) makes use of those pre-trained models to ease the feature learning process for other target domains that contain a smaller amount of training data. Currently, there are numerous ways to utilize features generated by transfer learning. Pre-trained CNN models prepare mid-/high-level features to work for different targeting problem domains. In this paper, a DL feature and model selection framework based on evolutionary programming is proposed to solve the challenges in visual data classification. It automates the process of discovering and obtaining the most representative features generated by the pre-trained DL models for different classification tasks. 

   more » « less
2. Contextually Guided Convolutional Neural Networks for Learning Most Transferable Representations

   https://doi.org/10.1109/ISM55400.2022.00047  

   Kursun, Olcay; Dinc, Semih; Favorov, Oleg V. (December 2022, 2022 IEEE International Symposium on Multimedia (ISM))

   Implementing local contextual guidance principles in a single-layer CNN architecture, we propose an efficient algorithm for developing broad-purpose representations (i.e., representations transferable to new tasks without additional training) in shallow CNNs trained on limited-size datasets. A contextually guided CNN (CG-CNN) is trained on groups of neighboring image patches picked at random image locations in the dataset. Such neighboring patches are likely to have a common context and therefore are treated for the purposes of training as belonging to the same class. Across multiple iterations of such training on different context-sharing groups of image patches, CNN features that are optimized in one iteration are then transferred to the next iteration for further optimization, etc. In this process, CNN features acquire higher pluripotency, or inferential utility for any arbitrary classification task. In our applications to natural images and hyperspectral images, we find that CG-CNN can learn transferable features similar to those learned by the first layers of the well-known deep networks and produce favorable classification accuracies. 

   more » « less
3. Taming the data deluge: A novel end‐to‐end deep learning system for classifying marine biological and environmental images

   https://doi.org/10.1002/lom3.10591  

   Bi, Hongsheng; Cheng, Yunhao; Cheng, Xuemin; Benfield, Mark C; Kimmel, David G; Zheng, Haiyong; Groves, Sabrina; Ying, Kezhen (January 2024, Limnology and Oceanography: Methods)

   Abstract Underwater imaging enables nondestructive plankton sampling at frequencies, durations, and resolutions unattainable by traditional methods. These systems necessitate automated processes to identify organisms efficiently. Early underwater image processing used a standard approach: binarizing images to segment targets, then integrating deep learning models for classification. While intuitive, this infrastructure has limitations in handling high concentrations of biotic and abiotic particles, rapid changes in dominant taxa, and highly variable target sizes. To address these challenges, we introduce a new framework that starts with a scene classifier to capture large within‐image variation, such as disparities in the layout of particles and dominant taxa. After scene classification, scene‐specific Mask regional convolutional neural network (Mask R‐CNN) models are trained to separate target objects into different groups. The procedure allows information to be extracted from different image types, while minimizing potential bias for commonly occurring features. Using in situ coastal plankton images, we compared the scene‐specific models to the Mask R‐CNN model encompassing all scene categories as a single full model. Results showed that the scene‐specific approach outperformed the full model by achieving a 20% accuracy improvement in complex noisy images. The full model yielded counts that were up to 78% lower than those enumerated by the scene‐specific model for some small‐sized plankton groups. We further tested the framework on images from a benthic video camera and an imaging sonar system with good results. The integration of scene classification, which groups similar images together, can improve the accuracy of detection and classification for complex marine biological images. 

   more » « less
4. Convolutional neural network learning for generic data classification

   Han, Huimei; Li, Ying; Zhu, Xingquan (March 2019, Information sciences)

   Convolutional Neural Network (CNN) uses convolutional layers to explore spatial/temporal adjacency to construct new feature representations. So, CNN is commonly used for data with strong temporal/spatial correlations, but cannot be directly applied to generic learning tasks. In this paper, we propose to enable CNN for learning from generic data to improve classification accuracy. To take the full advantage of CNN’s feature learning power, we propose to convert each instance of the original dataset into a synthetic matrix/image format. To maximize the correlation in the constructed matrix/image, we use 0/1 optimization to reorder features and ensure that the ones with strong correlations are adjacent to each other. By using a feature reordering matrix, we are able to create a synthetic image to represent each instance. Because the constructed synthetic image preserves the original feature values and correlation, CNN can be applied to learn effective features for classification. Experiments and comparisons, on 22 benchmark datasets, demonstrate clear performance gain of applying CNN to generic datasets, compared to conventional machine learning methods. Furthermore, our method consistently outperforms approaches which directly apply CNN to generic datasets in naive ways. This research allows deep learning to be broadly applied to generic datasets. 

   more » « less
5. Evaluating Multicultural Autism Screening for Toddlers Using Machine Learning on the QCHAT-10

   https://doi.org/10.1101/2024.11.12.24317211  

   Sollis, Lydia J; Wall, Dennis P; Washington, Peter Y (November 2024, medRxiv)

   ABSTRACT Early identification and intervention often leads to improved life outcomes for individuals with Autism Spectrum Disorder (ASD). However, traditional diagnostic methods are time-consuming, frequently delaying treatment. This study examines the application of machine learning (ML) techniques to 10-question Quantitative Checklist for Autism in Toddlers (QCHAT-10) datasets, aiming to evaluate the predictive value of questionnaire features and overall accuracy metrics across different cultures. We trained models using three distinct datasets from three different countries: Poland, New Zealand, and Saudi Arabia. The New Zealand and Saudi Arabian-trained models were both tested on the Polish dataset, which consisted of diagnostic class labels derived from clinical diagnostic processes. The Decision Tree, Random Forest, and XGBoost models were evaluated, with XGBoost consistently performing best. Feature importance rankings revealed little consistency across models; however, Recursive Feature Elimination (RFE) to select the models with the four most predictive features retained three common features. Both models performed similarly on the Polish test dataset with clinical diagnostic labels, with the New Zealand models with all 13 features achieving an AUROC of 0.94 ± 0.06, and the Saudi Model having an AUROC of 93% ± 6. This compared favorably to the cross-validation analysis of a Polish-trained model, which had an AUROC of 94% ± 5, suggesting that answers to the QCHAT-10 can be predictive of an official autism diagnosis, even across cultures. The New Zealand model with four features had an AUROC of 85% ± 13, and the Saudi model had a similar result of 87% ± 11. These results were somewhat lower than the Polish cross-validation AUROC of 91% ± 5. Adjusting probability thresholds improved sensitivity in some models, which is crucial for screening tools. However, this threshold adjustment often resulted in low levels of specificity during the final testing phase. Our findings suggest that these screening tools may generalize well across cultures; however, more research is needed regarding differences in feature importance for different populations. 

   more » « less