Journal: bioRxiv

Article Title: A computational probe into the behavioral and neural markers of atypical facial emotion processing in autism

doi: 10.1101/2021.03.24.436640

Figure Lengend Snippet: A. ANN models of the primate ventral stream (typically comprising V1, V2, V4 and IT like layers) can be trained to predict human facial emotion judgments. This involves building a regression model, i.e., determining the weights based on the model layer activations (as the predictor) to predict the image ground truth (“level of happiness”) on a set of training images, and then testing the predictions of this model on held-out images. B. An ANN model’s predicted psychometric curves (e.g., AlexNet, shown here) show the proportion of trials judged as “happy” as a function of facial emotion morph levels ranging from 0% happy (100% fearful; left) to 100% happy (0% fearful; right). This curve demonstrates that activations of ANN layers (layer ‘fc7’ that corresponds to the “model-IT” layer) can be successfully trained to predict facial emotions. C. Comparison of ANN’s image-level behavioral patterns with the behavior measured in Controls (x-axis) and IwA (y-axis). Four ANNs (with 5 models each generated from different layers of the ANNs are shown here in different colors. ANN predictions better match the behavior measured in the Controls compared to IwA. The correlation values (x and y axes) were corrected by the noise estimates per human population so that the differences are not due to differences in noise-levels in measurements across the IwA and Control subject pools. The dot size refers to the degree of discrepancy between ANN predictivity of Controls vs. IwA. D . A comparison of the ANN predictivity (results from AlexNet shown here) of behavior measured in IwA vs. Controls as function of model layers (convolutional (cnv) layers 1,3,4, and 5 and the fully connected layer 7, ‘fc7’ -- that approximately corresponds to the ventral stream cortical hierarchy). The difference between the ANN’s predictivity of behavior in IwA and Controls increases with depth and is referred to as Δ . E. Discriminability index (d’; ability to discriminate between image-level behavioral patterns measured in IwA vs. Controls ; see Methods) as a function of model layers (all four tested models shown separately in individual panels). The difference in ANN predictivity between Controls and IwA was largest at the deeper (more IT-like) layers of the models instead of earlier (more V1, V2, and V4-like) layers. Errorbars denote bootstrap confidence intervals. Facial images shown in this figure are morphed and processed version of the original face images. These images have full re-use permission.

Article Snippet: The model features, per layer, were extracted using the MATLAB command activations for AlexNet , VGGFace and EmotionNet in MATLAB-R 2020b.

Techniques: Generated

Journal: Diagnostics

Article Title: A Survey on Deep-Learning-Based Diabetic Retinopathy Classification

doi: 10.3390/diagnostics13030345

Figure Lengend Snippet: Retinopathy-grading-based studies during the period 2017–2020.

Article Snippet: Wang et al. [ ] , 2018 , AlexNet, VGG-16, and InceptionV3 , Kaggle EyePACS.

Techniques:

Journal: Scientific Reports

Article Title: Bayesian optimized multimodal deep hybrid learning approach for tomato leaf disease classification

doi: 10.1038/s41598-024-72237-x

Figure Lengend Snippet: Review of existing leaf disease methodologies with limitations.

Article Snippet: Da Rocha et al. , BO DL , AlexNet, ResNet50, SqueezeNet , Lack of extensive hyperparameter optimization.

Techniques: Extraction, Modification

Journal: Scientific Reports

Article Title: Bayesian optimized multimodal deep hybrid learning approach for tomato leaf disease classification

doi: 10.1038/s41598-024-72237-x

Figure Lengend Snippet: Comparison of the proposed approach with the latest approaches (tomato leaf 10 classes).

Article Snippet: Da Rocha et al. , BO DL , AlexNet, ResNet50, SqueezeNet , Lack of extensive hyperparameter optimization.

Techniques: Comparison, Modification

Journal: Scientific Reports

Article Title: Bayesian optimized multimodal deep hybrid learning approach for tomato leaf disease classification

doi: 10.1038/s41598-024-72237-x

Figure Lengend Snippet: Comparison of the suggested approach with recently established models for various crops.

Article Snippet: Da Rocha et al. , BO DL , AlexNet, ResNet50, SqueezeNet , Lack of extensive hyperparameter optimization.

Techniques: Comparison

Journal: Scientific Reports

Article Title: Bayesian optimized multimodal deep hybrid learning approach for tomato leaf disease classification

doi: 10.1038/s41598-024-72237-x

Figure Lengend Snippet: Comparison of the proposed model's training parameters with state-of-the-art models.

Article Snippet: Da Rocha et al. , BO DL , AlexNet, ResNet50, SqueezeNet , Lack of extensive hyperparameter optimization.

Techniques: Comparison

Journal: The Journal of Supercomputing

Article Title: An overview of machine learning methods in enabling IoMT-based epileptic seizure detection

doi: 10.1007/s11227-023-05299-9

Figure Lengend Snippet: A synopsis of DL techniques used in epilepsy detection automation

Article Snippet: [ ] , 2020 , Bonn , AlexNet , Softmax , Accuracy, sensitivity, specificity , 98.5%, 100%, 97.83%.

Techniques:

Journal: Scientific Reports

Article Title: Localization and Classification of Paddy Field Pests using a Saliency Map and Deep Convolutional Neural Network

doi: 10.1038/srep20410

Figure Lengend Snippet: The errors were computed over 3 epochs, of which each has 20000 iterations. Both learning processes used the same local receptive fields and drop ratio. The two loss functions were associated with a large L2-norm weight decay constant 0.005 (larger than that used in AlexNet ), which has proved to be useful for improving generalization of neural networks . Under these settings, softmax and hinge respectively achieved 0.932 and 0.891 in validation accuracy.

Article Snippet: AlexNet , Softmax , 0.834.

Techniques: Biomarker Discovery

Journal: Scientific Reports

Article Title: Localization and Classification of Paddy Field Pests using a Saliency Map and Deep Convolutional Neural Network

doi: 10.1038/srep20410

Figure Lengend Snippet: Comparison of DCNNs with other methods on the same dataset.

Article Snippet: AlexNet , Softmax , 0.834.

Techniques: Comparison

Journal: Sensors (Basel, Switzerland)

Article Title: ResNet Based Deep Features and Random Forest Classifier for Diabetic Retinopathy Detection †

doi: 10.3390/s21113883

Figure Lengend Snippet: Summary of related work.

Article Snippet: 2017, Mansour et al. [ ] , AlexNet with multiple optimization techniques , Accuracy of 95.26% with principal component analysis and 97.93% with FC7 features , EyePACS.

Techniques: Biomarker Discovery, Activation Assay