Related papers: Measurement-Adapted Eigentask Representations for …

Astronomers working with faint targets will benefit greatly from improved image quality on current and planned ground-based telescopes. At present, most adaptive optic systems are targeted at the highest resolution with bright guide stars.…

We consider the problem of subspace estimation in situations where the number of available snapshots and the observation dimension are comparable in magnitude. In this context, traditional subspace methods tend to fail because the…

We demonstrate accurate phase measurement from low photon level interference data using a constrained optimization method that takes into account the expected redundancy in the unknown phase function. This approach is shown to have…

A universal energy eigenvalue equation is proposed in this paper. It is proven that the unique set of eigenfunctions or preferred basis exists for any non-isolated sub-system. Applying the new eigenvalue equation to the relative motion of a…

Few-shot audio classification is an emerging topic that attracts more and more attention from the research community. Most existing work ignores the specificity of the form of the audio spectrogram and focuses largely on the embedding space…

Physical realizations of the canonical phase measurement for the optical phase are unknown. Single-shot phase estimation, which aims to determine the phase of an optical field in a single shot, is critical in quantum information processing…

Underwater target localization uses real-time sensory measurements to estimate the position of underwater objects of interest, providing critical feedback information for underwater robots. While acoustic sensing is the most acknowledged…

Recent advances in large-scale vision-language models have achieved impressive performance in various zero-shot image classification tasks. While prior studies have demonstrated significant improvements by introducing few-shot labelled…

Optical absorption measurements characterize a wide variety of systems from atomic gases to \emph{in-vivo} diagnostics of living organisms. Here we study the potential of non-classical techniques to reduce statistical noise below the…

Labeling data is often expensive and time-consuming, especially for tasks such as object detection and instance segmentation, which require dense labeling of the image. While few-shot object detection is about training a model on novel…

Recently, Motes et al. proposed in Phys. Rev. Lett. 114, 170802 (2015) a linear optics interferometer with N identical single photon input states as a tool for sub-shot-noise phase estimation which does not require NOON states sources. This…

Feature selection of high-dimensional labeled data with limited observations is critical for making powerful predictive modeling accessible, scalable, and interpretable for domain experts. Spectroscopy data, which records the interaction…

Absorption spectroscopy is routinely used to characterise chemical and biological samples. For the state-of-the-art in absorption spectroscopy, precision is theoretically limited by shot-noise due to the fundamental Poisson-distribution of…

In this paper, we seek to develop a versatile test-time adaptation (TTA) objective for a variety of tasks - classification and regression across image-, object-, and pixel-level predictions. We achieve this through a self-bootstrapping…

In microscopy image cell segmentation, it is common to train a deep neural network on source data, containing different types of microscopy images, and then fine-tune it using a support set comprising a few randomly selected and annotated…

Any-shot image classification allows to recognize novel classes with only a few or even zero samples. For the task of zero-shot learning, visual attributes have been shown to play an important role, while in the few-shot regime, the effect…

Few-shot learning is a relatively new technique that specializes in problems where we have little amounts of data. The goal of these methods is to classify categories that have not been seen before with just a handful of samples. Recent…

Prototype is widely used to represent internal structure of category for few-shot learning, which was proposed as a simple inductive bias to address the issue of overfitting. However, since prototype representation is normally averaged from…

Unsupervised node representations learnt using contrastive learning-based methods have shown good performance on downstream tasks. However, these methods rely on augmentations that mimic low-pass filters, limiting their performance on tasks…

We show that the way inference is performed in few-shot segmentation tasks has a substantial effect on performances -- an aspect often overlooked in the literature in favor of the meta-learning paradigm. We introduce a transductive…