Global ETD Search
Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.
Results
Showing 1 to 20 of 48 for “"Mixup"”.
-
MixUp as Directional Adversarial Training: A Unifying Understanding of MixUp and Adversarial Training
… a state-of-the-art regularization scheme called MixUp, and by contributing to the conceptual understanding of MixUp. MixUp is a data augmentation scheme in which pairs of training samples and their corresponding labels are mixed using linear coefficients. Without label mixing, MixUp becomes a …
-
On Mixup Training of Neural Networks
… a simple yet effective regularization scheme, Mixup, which has been proposed recently. Briefly speaking, Mixup creates synthetic examples by linearly interpolating random pairs of the real examples and uses the synthetic examples for training. Although Mixup has been empirically shown to be …
-
EXPLORING NEURAL NETWORK DEFENSES WITH ADVERSARIAL MIXUP
Neural networks (NNs) are vulnerable to adversarial examples, and extensive research is aimed at detecting them. However, detecting adversarial examples is not easy, even with the construction of new loss functions in a network. In this study, we introduce the Adversarial Mix up (AdvMix) network, a …
-
Text-Supervised Local Feature Mixup Towards Long-Tailed Image Categorization
… Furthermore, we propose a novel local feature mixup technique that takes advantage of the semantic relations between classes recognized by the pre-trained text encoder to further help alleviate the long-tailed problem. Our empirical study on several benchmark long-tailed tasks demonstrates the …
-
Resource-Efficient Machine Learning with Limited Data and Computation
… innovative graph data augmentation technique, G-Mixup, which mixes graph generators, e.g., graphon, to create synthetic training graphs. (ii) facing restricted computational resources, we developed an efficient graph neural network training method, MLPInit, which utilizes pre-trained MLP weights …
-
Analysis of Cellular and Subcellular Morphology using Machine Learning in Microscopy Images
… us ing data augmentation approaches, namely mixup, WGAN-div and novel nonlinear mixup, along with a minority class focussed sampling strategy. Classification performance evaluated using F1-score shows that the proposed deep learning framework out performs state-of-the art approaches on …
-
Duomenų augmentacijų efektyvumas nustatant ligas plaučių echoskopijose /
… ratio), affine (classical), and mixing (MixUp, CutMix, HMix, GMix) augmentations were reviewed and tested. Aſter training the models with augmentations, the resulting metrics were compared to the baseline augmentation-free model to assess the effect of augmentations on the metrics. COVID …
-
Exploring representation-level augmentation and RAG-based vulnerability augmentation with LLMs for vulnerability detection
… Specifically, we first explore traditional mixup-inspired augmentation methods that work at the representation level and show that these methods can be useful, although they cannot beat random oversampling. One possible reason is that mixing samples heavily degrades the integrity of the …
-
Gyvūnų aptikimas keliuose /
… animal detection accuracy, I implemented image mixup data augmentation to use while training the models. By using image mixup with beta distribution B(1, 1) the model reached 8.02% accuracy improvement. Additionally, I used brightness, contrast, hue, saturation and horizontal image flip image …
-
Towards Trustworthy Deep Learning and Large Language Models
… For robustness and generalization, we propose S-Mixup, a novel augmentation technique for graphs. Unlike standard Mixup, which assumes grid-like input, S-Mixup computes soft node alignments to enable meaningful interpolation between structurally diverse graphs. This improves generalization and …
-
Towards Trustworthy Deep Learning and Large Language Models
… For robustness and generalization, we propose S-Mixup, a novel augmentation technique for graphs. Unlike standard Mixup, which assumes grid-like input, S-Mixup computes soft node alignments to enable meaningful interpolation between structurally diverse graphs. This improves generalization and …
-
Towards Trustworthy Deep Learning and Large Language Models
… For robustness and generalization, we propose S-Mixup, a novel augmentation technique for graphs. Unlike standard Mixup, which assumes grid-like input, S-Mixup computes soft node alignments to enable meaningful interpolation between structurally diverse graphs. This improves generalization and …
-
Towards Trustworthy Deep Learning and Large Language Models
… For robustness and generalization, we propose S-Mixup, a novel augmentation technique for graphs. Unlike standard Mixup, which assumes grid-like input, S-Mixup computes soft node alignments to enable meaningful interpolation between structurally diverse graphs. This improves generalization and …
-
Towards Trustworthy Deep Learning and Large Language Models
… For robustness and generalization, we propose S-Mixup, a novel augmentation technique for graphs. Unlike standard Mixup, which assumes grid-like input, S-Mixup computes soft node alignments to enable meaningful interpolation between structurally diverse graphs. This improves generalization and …
-
Towards Trustworthy Deep Learning and Large Language Models
… For robustness and generalization, we propose S-Mixup, a novel augmentation technique for graphs. Unlike standard Mixup, which assumes grid-like input, S-Mixup computes soft node alignments to enable meaningful interpolation between structurally diverse graphs. This improves generalization and …
-
Learning Robust Graph Neural Networks with Limited Supervision
… a graph augmentation method called Graph Dual Mixup (GDM) is proposed for graph classification with scarce labels. GDM employs a graph structural auto-encoder to learn structural embeddings, applying mixup in the learned structural embedding space to generate new graph structures. Additionally, …
-
LEARNING MORE FROM LESS: RESOURCE-CONSTRAINED GENERATIVE AI FOR GENERATION AND UNDERSTANDING
… content. First, Few-shot Learning with hard-Mixup (FeLMi) addresses few-shot learning under data-constrained regimes through augmentation-driven, uncertainty-guided mixup strategies, improving robustness and generalization when labeled data is scarce. Building on this foundation, Cap2Aug …
-
Cross-Domain Imitation Learning from Human Videos for Robotic Manipulation
… Dynamic Time Warping trajectory mapping with MixUp interpolation in latent and action space. Evaluated on four real-world manipulation tasks across four robotic embodiments, ImMimic consistently improves task success rates and execution smoothness over robot-only and standard co-training …
-
Dataset Augmentation for Aspect Level Sentiment Analysis
… task in natural language processing. Manifold mixup is a data augmentation technique for images developed by Verma et al. to act as a regularizer for neural network and SwitchOut is another technique developed by Wang et al. to boost the performance of neural networks used for machine …
Page 1 of 3