Search

The horseshoe prior, a widely used handy alternative to the spike-and-slab prior, has proven to be an exceptional default global-local shrinkage prior in Bayesian inference and machine learning. However, designing tests with frequentist false discovery rate (FDR) control using the horseshoe prior or the general class of global-local shrinkage priors remains an open problem. In this paper, we propose a frequentist-assisted horseshoe procedure that not only resolves this long-standing FDR control issue for the high dimensional normal means testing problem but also exhibits satisfactory finite-sample FDR control under any desired nominal level for both large-scale multiple independent and correlated tests. We carry out the frequentist-assisted horseshoe procedure in an easy and intuitive way by using the minimax estimator of the global parameter of the horseshoe prior while maintaining the remaining full Bayes vanilla horseshoe structure. The results of both intensive simulations under different sparsity levels, and real-world data demonstrate that the frequentist-assisted horseshoe procedure consistently achieves robust finite-sample FDR control. Existing frequentist or Bayesian FDR control procedures can lose finite-sample FDR control in a variety of common sparse cases. Based on the intimate relationship between the minimax estimation and the level of FDR control discovered in this work, we point out potential generalizations to achieve FDR control for both more complicated models and the general global-local shrinkage prior family.

X-ray free electron lasers (XFELs) are the new generation of particle accelerator-based light sources, capable of producing tunable, high-power X-ray pulses that are increasingly vital across various scientific disciplines. Recently, continuous-wave (CW) XFELs driven by superconducting linear accelerators have garnered significant attention due to their ability to enhance availability by supporting multiple undulator lines simultaneously. However, different undulator lines typically require distinct electron beam qualities, particularly varying electron beam energy to achieve a wide range of photon energy tunability. Consequently, precise bunch-to-bunch control of electron beam energy is essential. A double-bend achromat based electron beam delay system has been proposed to enable multi-beam energy operations in CW-XFELs. In this paper, we introduce a novel delay system comprising four triple-bend achromats (TBAs). Based on parameters of the Shanghai High-Repetition-Rate XFEL and Extreme Light Facility, start-to-end simulations demonstrate that the TBA-based delay system achieves better electron beam qualities while providing a wide beam energy tuning range.

Recent advancement in 2D image diffusion models has driven significant progress in text-guided texture synthesis, enabling realistic, high-quality texture generation from arbitrary text prompts. However, current methods usually focus on synthesizing texture for single static 3D objects, and struggle to handle entire families of shapes, such as those produced by procedural programs. Applying existing methods naively to each procedural shape is too slow to support exploring different parameter settings at interactive rates, and also results in inconsistent textures across the procedural shapes. To this end, we introduce ProcTex, the first text-to-texture system designed for procedural 3D models. ProcTex enables consistent and real-time text-guided texture synthesis for families of shapes, which integrates seamlessly with the interactive design flow of procedural models. To ensure consistency, our core approach is to generate texture for the shape produced by one setting of the procedural parameters, followed by a texture transfer stage to apply the texture to other parameter settings. We also develop several techniques, including a novel UV displacement network for real-time texture transfer, the retexturing pipeline to support structural changes from discrete procedural parameters, and part-level UV texture map generation for local appearance editing. Extensive experiments on a diverse set of professional procedural models validate ProcTex's ability to produce high-quality, visually consistent textures while supporting real-time, interactive applications.

Neural Radiance Fields (NeRFs) have shown remarkable success in synthesizing photorealistic views from multi-view images of static scenes, but face challenges in dynamic, real-world environments with distractors like moving objects, shadows, and lighting changes. Existing methods manage controlled environments and low occlusion ratios but fall short in render quality, especially under high occlusion scenarios. In this paper, we introduce NeRF On-the-go, a simple yet effective approach that enables the robust synthesis of novel views in complex, in-the-wild scenes from only casually captured image sequences. Delving into uncertainty, our method not only efficiently eliminates distractors, even when they are predominant in captures, but also achieves a notably faster convergence speed. Through comprehensive experiments on various scenes, our method demonstrates a significant improvement over state-of-the-art techniques. This advancement opens new avenues for NeRF in diverse and dynamic real-world applications., Comment: CVPR 2024, first two authors contributed equally. Project Page: https://rwn17.github.io/nerf-on-the-go/

In electronic health records (EHR) analysis, clustering patients according to patterns in their data is crucial for uncovering new subtypes of diseases. Existing medical literature often relies on classical hypothesis testing methods to test for differences in means between these clusters. Due to selection bias induced by clustering algorithms, the implementation of these classical methods on post-clustering data often leads to an inflated type-I error. In this paper, we introduce a new statistical approach that adjusts for this bias when analyzing data collected over time. Our method extends classical selective inference methods for cross-sectional data to longitudinal data. We provide theoretical guarantees for our approach with upper bounds on the selective type-I and type-II errors. We apply the method to simulated data and real-world Acute Kidney Injury (AKI) EHR datasets, thereby illustrating the advantages of our approach.

The X-ray free-electron lasers (XFELs) are cutting-edge instruments pivotal in a broad range of fields, providing high-power X-ray pulses with durations spanning from femtoseconds to attoseconds. One of the critical challenges in XFEL facilities is the simultaneous accommodation of diverse requirements for XFEL operation modes and photon properties across different undulator lines. This paper proposes a dipole-kicker combination in the bunch compressors to vary the electron bunch length for the continuous-wave XFEL facilities driven by a superconducting linac. This method enables optimization of the electron bunch length on a per-bunch basis, tailored to each specific needs of each undulator. Through start-to-end simulations based on the parameters of the Shanghai high-repetition-rate XFEL and extreme light facility, we demonstrate the feasibility of this technique. The results show its effectiveness in enabling simultaneous operations of self-amplified spontaneous emission and externally seeded FEL across different undulator lines, ensuring optimal electron bunch compression for each undulator line.

We study the problem of community recovery and detection in multi-layer stochastic block models, focusing on the critical network density threshold for consistent community structure inference. Using a prototypical two-block model, we reveal a computational barrier for such multi-layer stochastic block models that does not exist for its single-layer counterpart: When there are no computational constraints, the density threshold depends linearly on the number of layers. However, when restricted to polynomial-time algorithms, the density threshold scales with the square root of the number of layers, assuming correctness of a low-degree polynomial hardness conjecture. Our results provide a nearly complete picture of the optimal inference in multiple-layer stochastic block models and partially settle the open question in Lei and Lin (2022) regarding the optimality of the bias-adjusted spectral method., Comment: 31 pages

Neural implicit representations have recently become popular in simultaneous localization and mapping (SLAM), especially in dense visual SLAM. However, previous works in this direction either rely on RGB-D sensors, or require a separate monocular SLAM approach for camera tracking and do not produce high-fidelity dense 3D scene reconstruction. In this paper, we present NICER-SLAM, a dense RGB SLAM system that simultaneously optimizes for camera poses and a hierarchical neural implicit map representation, which also allows for high-quality novel view synthesis. To facilitate the optimization process for mapping, we integrate additional supervision signals including easy-to-obtain monocular geometric cues and optical flow, and also introduce a simple warping loss to further enforce geometry consistency. Moreover, to further boost performance in complicated indoor scenes, we also propose a local adaptive transformation from signed distance functions (SDFs) to density in the volume rendering equation. On both synthetic and real-world datasets we demonstrate strong performance in dense mapping, tracking, and novel view synthesis, even competitive with recent RGB-D SLAM systems., Comment: Video: https://youtu.be/tUXzqEZWg2w

X-ray free-electron lasers (XFELs) are powerful tools to explore and study nature for achieving remarkable advances. Generally, seeded FELs are ideal sources for supplying full coherent soft x-ray pulses. Benefiting from the high-frequency up-conversion efficiency, the cascading configuration with echo-enabled harmonic generation (EEHG) and high-gain harmonic generation (HGHG) holds promising prospects for generating full coherent radiation at 1 nm wavelength. In this paper, we design and optimize EEHG-HGHG configuration using parameters of Shanghai High-Repetition-Rate XFEL and Extreme Light Facility. In addition, we systematically analyze the effect of relative timing jitter on the output FEL performance based on various start-to-end electron beams. The intensive numerical simulations show that the cascaded EEHG-HGHG scheme can achieve 1 nm FEL pulses with peak power up to 15 GW. Further sensitivity analysis indicates that the relative timing jitter between the electron beam and seed laser has a significant impact on the FEL performance. The RMS timing jitter of 3 fs can lead to the final output pulse energy fluctuations of 29.16%., Comment: 21 pages, 12 figures

We propose a novel method to reconstruct the 3D shapes of transparent objects using hand-held captured images under natural light conditions. It combines the advantage of explicit mesh and multi-layer perceptron (MLP) network, a hybrid representation, to simplify the capture setting used in recent contributions. After obtaining an initial shape through the multi-view silhouettes, we introduce surface-based local MLPs to encode the vertex displacement field (VDF) for the reconstruction of surface details. The design of local MLPs allows to represent the VDF in a piece-wise manner using two layer MLP networks, which is beneficial to the optimization algorithm. Defining local MLPs on the surface instead of the volume also reduces the searching space. Such a hybrid representation enables us to relax the ray-pixel correspondences that represent the light path constraint to our designed ray-cell correspondences, which significantly simplifies the implementation of single-image based environment matting algorithm. We evaluate our representation and reconstruction algorithm on several transparent objects with ground truth models. Our experiments show that our method can produce high-quality reconstruction results superior to state-of-the-art methods using a simplified data acquisition setup.

Neural implicit representations have recently shown encouraging results in various domains, including promising progress in simultaneous localization and mapping (SLAM). Nevertheless, existing methods produce over-smoothed scene reconstructions and have difficulty scaling up to large scenes. These limitations are mainly due to their simple fully-connected network architecture that does not incorporate local information in the observations. In this paper, we present NICE-SLAM, a dense SLAM system that incorporates multi-level local information by introducing a hierarchical scene representation. Optimizing this representation with pre-trained geometric priors enables detailed reconstruction on large indoor scenes. Compared to recent neural implicit SLAM systems, our approach is more scalable, efficient, and robust. Experiments on five challenging datasets demonstrate competitive results of NICE-SLAM in both mapping and tracking quality. Project page: https://pengsongyou.github.io/nice-slam, Comment: CVPR 2022, first two authors contributed equally. Project page: https://pengsongyou.github.io/nice-slam

The formation of a double-horn current profile is a challenging issue in the achievement of electron bunch with high peak current, especially for high-repetition-rate X-ray free-electron lasers (XFELs) where more nonlinear and intricate beam dynamics propagation is involved. Here, we propose to correct the nonlinear beam longitudinal phase space distortion in the photoinjector section with a dual-mode buncher. In combination with the evolutionary many-objective beam dynamics optimization, this method is shown to be effective in manipulating the longitudinal phase space at the injector exit. Furthermore, the formation of the current spike is avoided after the multi-stage charge density modulation and electron bunch with a peak current of 1.6 kA is achieved for 100-pC bunch charge.Start-to-end simulations based on the Shanghai high-repetition-rate XFEL and extreme light facility demonstrate that the proposed scheme can increase the FEL pulse energy by more than 3 times in a cascading operation of echo-enabled harmonic generation and high-gain harmonic generation. Moreover, this method can also be used for longitudinal phase space shaping of electron beams operating at a high repetition rate to meet the specific demands of different researches.

SHINE, as the first hard x-ray free-electron-laser (FEL) facility in China, is designed to provide high-brightness FEL lasing under high-repetition-rate operation mode. In order to drive x-ray FEL pulses with high qualities, the photoinjector section is deployed to provide the specified electron beam with low transverse emittance and high brightness. Normally the multi-objective optimization algorithm is employed in the injector beam dynamics design. In this paper, the many-objective optimization algorithm NSGA-III is introduced to the injector physical design for optimizing the 4 beam quality properties for the first time. The results of the optimization are presented and analyzed. The application of NSGA-III can provide guidance for further physical research as well as improve the beam dynamics optimization efficiency.

Purpose: Acquisition timing and B$_1$ calibration are two key factors that affect the quality and accuracy of hyperpolarized $^{13}$C MRI. The goal of this project was to develop a new approach using regional bolus tracking to trigger Bloch-Siegert B$_1$ mapping and real-time B$_1$ calibration based on regional B$_1$ measurements, followed by dynamic imaging of hyperpolarized $^{13}C$ metabolites in vivo. Methods: The proposed approach was implemented on a system which allows real-time data processing and real-time control on the sequence. Real-time center frequency calibration upon the bolus arrival was also added. The feasibility of applying the proposed framework for in vivo hyperpolarized $^{13}$C imaging was tested on healthy rats, tumor-bearing mice and a healthy volunteer on a clinical 3T scanner following hyperpolarized [1-$^{13}$C]pyruvate injection. Multichannel receive coils were used in the human study. Results: Automatic acquisition timing based on either regional bolus peak or bolus arrival was achieved with the proposed framework. Reduced blurring artifacts in real-time reconstructed images were observed with real-time center frequency calibration. Real-time computed B$_1$ scaling factors agreed with real-time acquired B$_1$ maps. Flip angle correction using B$_1$ maps results in a more consistent quantification of metabolic activity (i.e, pyruvate-to-lactate conversion, k$_{PL}$). Experiment recordings are provided to demonstrate the real-time actions during the experiment. Conclusion: The proposed method was successfully demonstrated on animals and a human volunteer, and is anticipated to improve the efficient use of the hyperpolarized signal as well as the accuracy and robustness of hyperpolarized $^{13}$C imaging.