Related papers: Pretraining Objective Matters in Extreme Low-Data …
Can pretrained models generalize to new datasets without any retraining? We deploy pretrained image models on datasets they were not trained for, and investigate whether their embeddings form meaningful clusters. Our suite of benchmarking…
Accurate and robust medical image classification is paramount for early disease diagnosis and treatment planning. However, challenges such as limited annotated data, high intra-class variability, and subtle inter-class differences often…
Meta-learning is a powerful paradigm for tackling few-shot tasks. However, recent studies indicate that models trained with the whole-class training strategy can achieve comparable performance to those trained with meta-learning in few-shot…
Pre-training Graph Neural Networks (GNN) via self-supervised contrastive learning has recently drawn lots of attention. However, most existing works focus on node-level contrastive learning, which cannot capture global graph structure. The…
Dataset pruning reduces the storage and training costs of deep learning by selecting an informative subset from a large dataset. However, most existing pruning methods require fully labeled data, which limits their applicability in…
Few-shot learning is challenging due to its very limited data and labels. Recent studies in big transfer (BiT) show that few-shot learning can greatly benefit from pretraining on large scale labeled dataset in a different domain. This paper…
Ultrasound is widely used in obstetric care due to its safety, accessibility, and real-time imaging. However, interpretation remains operator-dependent and susceptible to noise and artifacts. Deep learning models have shown strong…
Addressing performance degradation in 3D LiDAR semantic segmentation due to domain shifts (e.g., sensor type, geographical location) is crucial for autonomous systems, yet manual annotation of target data is prohibitive. This study…
Background and objective: Expert annotations limit large-scale supervised pretraining in medical imaging, while ubiquitous metadata (modality, anatomical region) remain underused. We introduce ModAn-MulSupCon, a modality- and anatomy-aware…
Work in deep clustering focuses on finding a single partition of data. However, high-dimensional data, such as images, typically feature multiple interesting characteristics one could cluster over. For example, images of objects against a…
Fine-grained categories are more difficulty distinguished than generic categories due to the similarity of inter-class and the diversity of intra-class. Therefore, the fine-grained visual categorization (FGVC) is considered as one of…
Deep neural networks trained with standard cross-entropy loss memorize noisy labels, which degrades their performance. Most research to mitigate this memorization proposes new robust classification loss functions. Conversely, we propose a…
Multiclass neural network classifiers are typically trained using cross-entropy loss but evaluated using metrics derived from the confusion matrix, such as Accuracy, $F_\beta$-Score, and Matthews Correlation Coefficient. This mismatch…
Unsupervised visual representation learning has gained much attention from the computer vision community because of the recent achievement of contrastive learning. Most of the existing contrastive learning frameworks adopt the instance…
The majority of existing Unsupervised Domain Adaptation (UDA) methods presumes source and target domain data to be simultaneously available during training. Such an assumption may not hold in practice, as source data is often inaccessible…
Vector search systems built on frozen vision encoders face queries from unseen classes at deployment, yet existing adapter training collapses under this shift: high-capacity adapters with global contrastive losses silently reassign…
Deep neural networks are increasingly used in medical imaging for tasks such as pathological classification, but they face challenges due to the scarcity of high-quality, expert-labeled training data. Recent efforts have utilized…
Inspired by recent advances in diffusion models, which are reminiscent of denoising autoencoders, we investigate whether they can acquire discriminative representations for classification via generative pre-training. This paper shows that…
Advancements in open-source pre-trained backbones make it relatively easy to fine-tune a model for new tasks. However, this lowered entry barrier poses potential risks, e.g., bad actors developing models for harmful applications. A question…
Training deep networks with noisy labels leads to poor generalization and degraded accuracy due to overfitting to label noise. Existing approaches for learning with noisy labels often rely on the availability of a clean subset of data. By…