Journal: iScience

Article Title: Perceptual reorganization from prior knowledge emerges late in childhood

doi: 10.1016/j.isci.2024.108787

Figure Lengend Snippet: Comparison of CNN performance to human performance (A) Image recognition accuracy of CNNs (AlexNet, CORnet-S, NASNet-large; grayscale bars) and human performance (colored bars, error bars show bootstrapped 95% confidence intervals) for each condition (grayscale and two-tone) of images used in current study (novel Iiage set) and ImageNet images used to train CNN models (trained image set). CNN performance: percent of images correctly labeled; human performance: naive naming accuracy. (B) Imagewise comparison of mean 4- to 5-year-old naive naming accuracy and NASNet performance (measured by probability weighting of the highest-scoring correct label) for Catch and two-tone trials of novel image set. Red line shows linear least-squares line of best fit, dashed line shows x = y.

Article Snippet: Pretrained AlexNet and NASNet-large models were acquired from and run with Matlab R2020a’s Deep Learning Toolbox.

Techniques: Comparison, Labeling

Journal: Diagnostics

Article Title: EfficientNetV2 Based Ensemble Model for Quality Estimation of Diabetic Retinopathy Images from DeepDRiD

doi: 10.3390/diagnostics13040622

Figure Lengend Snippet: Previous works on assessing the fundus image quality using different machine learning and deep learning methods on various private and public datasets.

Article Snippet: Shi C et al. [ ] , Pretrained ResNet50 , Kaggle—2434 (2-class) , Multiple resolutions , Accuracy: 98.6 Sensitivity: 98.0 Specificity: 99.1.

Techniques:

Journal: Computational Intelligence and Neuroscience

Article Title: A Novel Hybrid Deep Learning Model for Metastatic Cancer Detection

doi: 10.1155/2022/8141530

Figure Lengend Snippet: Comparative study of the proposed model with recent ML/DL models.

Article Snippet: [ ] , Breast cancer , Pretrained CNN , Binary class , Kaggle , 90%.

Techniques: Diagnostic Assay