Search

Coreset selection is a method for selecting a small, representative subset of an entire dataset. It has been primarily researched in image classification, assuming there is only one object per image. However, coreset selection for object detection is more challenging as an image can contain multiple objects. As a result, much research has yet to be done on this topic. Therefore, we introduce a new approach, Coreset Selection for Object Detection (CSOD). CSOD generates imagewise and classwise representative feature vectors for multiple objects of the same class within each image. Subsequently, we adopt submodular optimization for considering both representativeness and diversity and utilize the representative vectors in the submodular optimization process to select a subset. When we evaluated CSOD on the Pascal VOC dataset, CSOD outperformed random selection by +6.4%p in AP$_{50}$ when selecting 200 images., Comment: Accepted by CVPR 2024: 1st Workshop on Dataset Distillation for Computer Vision

Recent advancements in the Neural Radiance Field (NeRF) have bolstered its capabilities for novel view synthesis, yet its reliance on dense multi-view training images poses a practical challenge. Addressing this, we propose HourglassNeRF, an effective regularization-based approach with a novel hourglass casting strategy. Our proposed hourglass is conceptualized as a bundle of additional rays within the area between the original input ray and its corresponding reflection ray, by featurizing the conical frustum via Integrated Positional Encoding (IPE). This design expands the coverage of unseen views and enables an adaptive high-frequency regularization based on target pixel photo-consistency. Furthermore, we propose luminance consistency regularization based on the Lambertian assumption, which is known to be effective for training a set of augmented rays under the few-shot setting. Leveraging the inherent property of a Lambertian surface, which retains consistent luminance irrespective of the viewing angle, we assume our proposed hourglass as a collection of flipped diffuse reflection rays and enhance the luminance consistency between the original input ray and its corresponding hourglass, resulting in more physically grounded training framework and performance improvement. Our HourglassNeRF outperforms its baseline and achieves competitive results on multiple benchmarks with sharply rendered fine details. The code will be available., Comment: 21 pages, 11 figures

The kernel trick in supervised learning signifies transformations of an inner product by a feature map, which then restructures training data in a larger Hilbert space according to an endowed inner product. A quantum feature map corresponds to an instance with a Hilbert space of quantum states by fueling quantum resources to machine learning algorithms. In this work, we point out that the quantum state space is specific such that a measurement postulate characterizes an inner product and that manipulation of quantum states prepared from classical data cannot enhance the distinguishability of data points. We present a feature map for quantum data as a probabilistic manipulation of quantum states to improve supervised learning algorithms., Comment: QCNC 2024 accepted paper (Main conference)

Finding the optimal model complexity that minimizes the generalization error (GE) is a key issue of machine learning. For the conventional supervised learning, this task typically involves the bias-variance tradeoff: lowering the bias by making the model more complex entails an increase in the variance. Meanwhile, little has been studied about whether the same tradeoff exists for unsupervised learning. In this study, we propose that unsupervised learning generally exhibits a two-component tradeoff of the GE, namely the model error and the data error -- using a more complex model reduces the model error at the cost of the data error, with the data error playing a more significant role for a smaller training dataset. This is corroborated by training the restricted Boltzmann machine to generate the configurations of the two-dimensional Ising model at a given temperature and the totally asymmetric simple exclusion process with given entry and exit rates. Our results also indicate that the optimal model tends to be more complex when the data to be learned are more complex., Comment: 15 pages, 7 figures

Multiple light scattering hampers imaging objects in complex scattering media. Approaches used in real practices mainly aim to filter out multiple scattering obscuring the ballistic waves that travel straight through the scattering medium. Here, we propose a method that makes the deterministic use of multiple scattering for microscopic imaging of an object embedded deep within scattering media. The proposed method finds a stack of multiple complex phase plates that generate similar light trajectories as the original scattering medium. By implementing the inverse scattering using the identified phase plates, our method rectifies multiple scattering and amplifies ballistic waves by almost 600 times, which leads to a substantial increase in imaging depth. Our study marks an important milestone in solving the longstanding high-order inverse scattering problems.

Recent end-to-end multi-object detectors simplify the inference pipeline by removing hand-crafted processes such as non-maximum suppression (NMS). However, during training, they still heavily rely on heuristics and hand-crafted processes which deteriorate the reliability of the predicted confidence score. In this paper, we propose a novel framework to train an end-to-end multi-object detector consisting of only two terms: negative log-likelihood (NLL) and a regularization term. In doing so, the multi-object detection problem is treated as density estimation of the ground truth bounding boxes utilizing a regularized mixture density model. The proposed \textit{end-to-end multi-object Detection with a Regularized Mixture Model} (D-RMM) is trained by minimizing the NLL with the proposed regularization term, maximum component maximization (MCM) loss, preventing duplicate predictions. Our method reduces the heuristics of the training process and improves the reliability of the predicted confidence score. Moreover, our D-RMM outperforms the previous end-to-end detectors on MS COCO dataset., Comment: Accepted at ICML 2023

The combustion delay mixture of tungsten (W), potassium perchlorate (KClO4), and barium chromate (BaCrO4), also known as the WKB mixture, has long been considered to be an integral part of military-grade ammunition. Despite its long history, however, their progressive reaction dynamics remains a question mark, especially due to the complex nature of their combustion reaction. As opposed to a one-step oxidation commonly observed in conventional combustions, the WKB mixture is associated with a multibody reaction between its solid-state components. To this end, the emergence of three combustion peaks, which we corresponded with disparate chemical reactions, was observed using thermogravimetric analysis on two separate WKB mixtures with differing mixture ratios. We applied the stepwise isoconversional method on each of the peaks to match the combustion chemistry it represents to the Sestak-Berggren model and computed the conceptual activation energy. Further plotting the logarithmic pre-exponential factor as a function of the reaction progress, we demonstrate a method of using the plot as an intuitive tool to understand the dynamics of individual reactions that compose multi-step chemical reactions. Our study provides a systematic approach in visualizing the reaction chemistry, thereby strengthening the analytical arsenal against reaction dynamics of combustion compounds in general., Comment: 10 pages, 5 figures, and 1 table

Few-shot object detection aims to detect instances of specific categories in a query image with only a handful of support samples. Although this takes less effort than obtaining enough annotated images for supervised object detection, it results in a far inferior performance compared to the conventional object detection methods. In this paper, we propose a meta-learning-based approach that considers the unique characteristics of each support sample. Rather than simply averaging the information of the support samples to generate a single prototype per category, our method can better utilize the information of each support sample by treating each support sample as an individual prototype. Specifically, we introduce two types of attention mechanisms for aggregating the query and support feature maps. The first is to refine the information of few-shot samples by extracting shared information between the support samples through attention. Second, each support sample is used as a class code to leverage the information by comparing similarities between each support feature and query features. Our proposed method is complementary to the previous methods, making it easy to plug and play for further improvement. We have evaluated our method on PASCAL VOC and COCO benchmarks, and the results verify the effectiveness of our method. In particular, the advantages of our method are maximized when there is more diversity among support data.

Deep-tissue optical imaging suffers from the reduction of resolving power due to tissue-induced optical aberrations and multiple scattering noise. Reflection matrix approaches recording the maps of backscattered waves for all the possible orthogonal input channels have provided formidable solutions for removing severe aberrations and recovering the ideal diffraction-limited spatial resolution without relying on fluorescence labeling and guide stars. However, measuring the full input-output response of the tissue specimen is time-consuming, making the real-time image acquisition difficult. Here, we present the use of a time-reversal matrix, instead of the reflection matrix, for fast high-resolution volumetric imaging of a mouse brain. The time-reversal matrix reduces two-way problem to one-way problem, which effectively relieves the requirement for the coverage of input channels. Using a newly developed aberration correction algorithm designed for the time-reversal matrix, we demonstrated the correction of complex aberrations using as small as 2 % of the complete basis while maintaining the image reconstruction fidelity comparable to the fully sampled reflection matrix. Due to nearly 100-fold reduction in the matrix recording time, we could achieve real-time aberration-correction imaging for a field of view of 40 x 40 microns (176 x 176 pixels) at a frame rate of 80 Hz. Furthermore, we demonstrated high-throughput volumetric adaptive optical imaging of a mouse brain by recording a volume of 128 x 128 x 125 microns (568 x 568 x 125 voxels) in 3.58 s, correcting tissue aberrations at each and every 1-micron depth section, and visualizing myelinated axons with a lateral resolution of 0.45 microns and an axial resolution of 2 microns.

In multi-object detection using neural networks, the fundamental problem is, "How should the network learn a variable number of bounding boxes in different input images?". Previous methods train a multi-object detection network through a procedure that directly assigns the ground truth bounding boxes to the specific locations of the network's output. However, this procedure makes the training of a multi-object detection network too heuristic and complicated. In this paper, we reformulate the multi-object detection task as a problem of density estimation of bounding boxes. Instead of assigning each ground truth to specific locations of network's output, we train a network by estimating the probability density of bounding boxes in an input image using a mixture model. For this purpose, we propose a novel network for object detection called Mixture Density Object Detector (MDOD), and the corresponding objective function for the density-estimation-based training. We applied MDOD to MS COCO dataset. Our proposed method not only deals with multi-object detection problems in a new approach, but also improves detection performances through MDOD. The code is available: https://github.com/yoojy31/MDOD., Comment: 10 pages, 7 figures

We present a laser scanning reflection-matrix microscopy combining the scanning of laser focus and the wide-field mapping of the electric field of the backscattered waves for eliminating higher-order aberrations even in the presence of strong multiple light scattering noise. Unlike conventional confocal laser scanning microscopy, we record the amplitude and phase maps of reflected waves from the sample not only at the confocal pinhole, but also at other non-confocal points. These additional measurements lead us to constructing a time-resolved reflection matrix, with which the sample-induced aberrations for the illumination and detection pathways are separately identified and corrected. We realized in vivo reflectance imaging of myelinated axons through an intact skull of a living mouse with the spatial resolution close to the ideal diffraction limit. Furthermore, we demonstrated near-diffraction-limited multiphoton imaging through an intact skull by physically correcting the aberrations identified from the reflection matrix. The proposed method is expected to extend the range of applications, where the knowledge of the detailed microscopic information deep within biological tissues is critical., Comment: 14 pages, 4 figures

Given an acyclic digraph $D$, the competition graph of $D$, denoted by $C(D)$, is the simple graph having vertex set $V(D)$ and edge set $\{uv \mid (u, w), (v, w) \in A(D) \text{ for some } w \in V(D) \}$. The phylogeny graph of an acyclic digraph $D$, denoted by $P(D)$, is the graph with the vertex set $V(D)$ and the edge set $E(U(D)) \cup E(C(D))$ where $U(D)$ denotes the underlying graph of $D$. The notion of phylogeny graphs was introduced by Roberts and Sheng~\cite{roberts1997phylogeny} as a variant of competition graph. Moral graphs having arisen from studying Bayesian networks are the same as phylogeny graphs. In this paper, we integrate the existing theorems computing phylogeny numbers of connected graph with a small number of triangles into one proposition: for a graph $G$ containing at most two triangle, $ |E(G)|-|V(G)|-2t(G)+d(G)+1 \le p(G) \le |E(G)|-|V(G)|-t(G)+1 $ where $t(G)$ and $d(G)$ denote the number of triangles and the number of diamond in $G$, respectively. Then we show that these inequalities hold for graphs with many triangles. In the process of showing it, we derive a useful theorem which plays a key role in deducing various meaningful results including a theorem that answers a question given by Wu~{\it et al.}~\cite{Wu2019}.

In this paper, we treat the image generation task using an autoencoder, a representative latent model. Unlike many studies regularizing the latent variable's distribution by assuming a manually specified prior, we approach the image generation task using an autoencoder by directly estimating the latent distribution. To this end, we introduce 'latent density estimator' which captures latent distribution explicitly and propose its structure. Through experiments, we show that our generative model generates images with the improved visual quality compared to previous autoencoder-based generative models.