TL;DR
The results showed model’s robustness against the type of salt-and-pepper noise, moderate behavior against the type of speckle and sensitivity to high Gaussian variance noise, and the range of noise variance and density is the best level for practical operation.
A new approach is presented here that integrates traditional handcrafted texture features such as Histogram of Oriented Gradients (HOG), Local Binary Patterns (LBP) and Gray-Level Co-occurrence Matrix (GLCM) with deep learning models with deep learning models.
An edge-aware line-level HTR framework that extends a CNN-Transformer baseline with a learnable edge-extraction channel and Squeeze-and-Excitation channel attention and shows that combining learnable structural cues with channel-wise attention has improved robustness for degradation-prone historical manuscript collections.
The Handwritten Digit Recognition System is a machine learning and deep learning–based project developed to accurately identify handwritten numerical digits from input images by using image processing techniques and a Convolutional Neural Network model trained on the MNIST dataset.
The efficient-KAN literature---covering Chebyshev, wavelet, and radial-basis-function variants of the original Kolmogorov-Arnold Network---has been benchmarked almost entirely on clean data. We show that this choice conceals a large capability difference between architectures: ChebyKAN's test MSE (evaluated against clean ground truth) increases by a factor of 10.6x when training data is corrupted with sigma=0.1 noise, versus 7.9x for vanilla KAN, 1.7x for a standard MLP, and just 1.4x for our proposed ER-KAN. ER-KAN combines three design choices targeting the noisy, data-scarce setting: shared Gaussian RBF bases across all edges in a layer (providing locality and efficient parameterisation), curriculum noise injection during training (explicitly teaching noise robustness), and entropy-weighted adaptive regularisation (preventing overfitting at small N). The result is a 595-parameter network that matches MLP accuracy at moderate noise while degrading far more gracefully as noise grows. We evaluate on eight analytic functions (N in {50, 200, 500}, sigma in {0, 0.03, 0.1}), on a damped harmonic oscillator physics-informed neural network where ER-KAN achieves 4.2x lower solution MSE than MLP, and on a Burgers'equation PINN where all models fail to converge---a genuine limitation we report rather than suppress. We introduce the noise degradation ratio as a simple complementary metric and recommend it become a standard reporting requirement for efficient-KAN papers.
A compact convolutional network for 46-class DHCD Devanagari recognition and reached 99.73%, the highest reported at 15.6x smaller than prior state-of-the-art, effectively reaching the saturation point.
We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.
