arXiv subjects

Yuxin Feng

Publications and source records attributed to Yuxin Feng.

8 recordsLinked to original sources

arXiv2026

Bilevel Layer-Positioning LoRA for Real Image Dehazing

Learning-based real image dehazing methods have achieved notable progress, yet they still face adaptation challenges in diverse real haze scenes. These challenges mainly stem from the lack of effective unsupervised mechanisms for unlabeled data and the heavy cost of full model fine-tuning. To address these challenges, we propose the haze-to-clear text-directed loss that leverages CLIP's cross-modal capabilities to reformulate real image dehazing as a semantic alignment problem in latent space, thereby providing explicit unsupervised cross-modal guidance in the absence of reference images. Furthermore, we introduce the Bilevel Layer-positioning LoRA (BiLaLoRA) strategy, which learns both the LoRA parameters and automatically search the injection layers, enabling targeted adaptation of critical network layers. Extensive experiments demonstrate our superiority against state-of-the-art methods on multiple real-world dehazing benchmarks. The code is publicly available at https://github.com/YanZhang-zy/BiLaLoRA.

cs.CV↗

arXiv2025

Benchmarking Endoscopic Surgical Image Restoration and Beyond

In endoscopic surgery, a clear and high-quality visual field is critical for surgeons to make accurate intraoperative decisions. However, persistent visual degradation, including smoke generated by energy devices, lens fogging from thermal gradients, and lens contamination due to blood or tissue fluid splashes during surgical procedures, severely impairs visual clarity. These degenerations can seriously hinder surgical workflow and pose risks to patient safety. To systematically investigate and address various forms of surgical scene degradation, we introduce a real- world open-source surgical image restoration dataset covering endoscopic environments, called SurgClean, which involves multi-type image restoration tasks from two medical sites, i.e., desmoking, defogging, and desplashing. SurgClean comprises 3,113 images with diverse degradation types and corresponding paired reference labels. Based on SurgClean, we establish a standardized evaluation benchmark and provide performance for 22 representative generic task-specific image restoration approaches, including 12 generic and 10 task-specific image restoration approaches. Experimental results reveal substantial performance gaps relative to clinical requirements, highlighting a critical opportunity for algorithm advancements in intelligent surgical restoration. Furthermore, we explore the degradation discrepancies between surgical and natural scenes from structural perception and semantic under- standing perspectives, providing fundamental insights for domain-specific image restoration research. Our work aims to empower restoration algorithms and improve the efficiency of clinical procedures.

cs.CV↗

arXiv2025

CoA: Towards Real Image Dehazing via Compression-and-Adaptation

Learning-based image dehazing algorithms have shown remarkable success in synthetic domains. However, real image dehazing is still in suspense due to computational resource constraints and the diversity of real-world scenes. Therefore, there is an urgent need for an algorithm that excels in both efficiency and adaptability to address real image dehazing effectively. This work proposes a Compression-and-Adaptation (CoA) computational flow to tackle these challenges from a divide-and-conquer perspective. First, model compression is performed in the synthetic domain to develop a compact dehazing parameter space, satisfying efficiency demands. Then, a bilevel adaptation in the real domain is introduced to be fearless in unknown real environments by aggregating the synthetic dehazing capabilities during the learning process. Leveraging a succinct design free from additional constraints, our CoA exhibits domain-irrelevant stability and model-agnostic flexibility, effectively bridging the model chasm between synthetic and real domains to further improve its practical utility. Extensive evaluations and analyses underscore the approach's superiority and effectiveness. The code is publicly available at https://github.com/fyxnl/COA.

cs.CV↗

arXiv2025

FrustumFusionNets: A Three-Dimensional Object Detection Network Based on Tractor Road Scene

To address the issues of the existing frustum-based methods' underutilization of image information in road three-dimensional object detection as well as the lack of research on agricultural scenes, we constructed an object detection dataset using an 80-line Light Detection And Ranging (LiDAR) and a camera in a complex tractor road scene and proposed a new network called FrustumFusionNets (FFNets). Initially, we utilize the results of image-based two-dimensional object detection to narrow down the search region in the three-dimensional space of the point cloud. Next, we introduce a Gaussian mask to enhance the point cloud information. Then, we extract the features from the frustum point cloud and the crop image using the point cloud feature extraction pipeline and the image feature extraction pipeline, respectively. Finally, we concatenate and fuse the data features from both modalities to achieve three-dimensional object detection. Experiments demonstrate that on the constructed test set of tractor road data, the FrustumFusionNetv2 achieves 82.28% and 95.68% accuracy in the three-dimensional object detection of the two main road objects, cars and people, respectively. This performance is 1.83% and 2.33% better than the original model. It offers a hybrid fusion-based multi-object, high-precision, real-time three-dimensional object detection technique for unmanned agricultural machines in tractor road scenarios. On the Karlsruhe Institute of Technology and Toyota Technological Institute (KITTI) Benchmark Suite validation set, the FrustumFusionNetv2 also demonstrates significant superiority in detecting road pedestrian objects compared with other frustum-based three-dimensional object detection methods.

cs.CV↗

arXiv2007

GRO J1655-40: from ASCA and XMM-Newton Observations

We have analysed four ASCA observations (1994--1995, 1996--1997) and three XMM-Newton observations (2005) of this source, in all of which the source is in high/soft state. We modeled the continuum spectra with relativistic disk model kerrbb, estimated the spin of the central black hole, and constrained the spectral hardening factor f_col and the distance. If kerrbb model applies, for normally used value of f_col, the distance cannot be very small, and f_col changes with observations.

astro-ph↗

arXiv2004

Studying the Properties of Accretion Disks and Coronae in Black Hole X-ray Binaries with Monte-Carlo Simulations

Understanding the properties of the hot corona is important for studying the accretion disks in black hole X-ray binary systems. Using the Monte-Carlo technique to simulate the inverse Compton scattering between photons emitted from the cold disk and electrons in the hot corona, we have produced two table models in the $XSPEC$ format for the spherical corona case and the disk-like (slab) corona case. All parameters in our table models are physical properties of the system and can be derived from data fitting directly. Applying the models to broad-band spectra of the black hole candidate XTE J2012+381 observed with BeppoSAX, we estimated the size of the corona and the inner radius of the disk. The size of the corona in this system is several tens of gravitational radius, and the substantial increase of the inner disk radius during the transit from hard-state to soft-state is not found.

astro-ph↗

arXiv2002

Relativistic Effects on the Appearance of a Clothed Black Hole

For an accretion disk around a black hole, the strong relativistic effects affect every aspect of the radiation from the disk, including its spectrum, light-curve, and image. This work investigates in detail how the images of a thin disk around a black hole will be distorted, and what the observer will see from different viewing angles and in different energy bands.

astro-ph↗

arXiv2001

A Peculiar Flaring Episode of Cygnus X-1

Recent monitoring of Cyg X-1 with {\em RXTE} revealed a period of intense flaring, which started in October of 2000 and lasted until March of 2001. The source exhibited some quite unusual behaviors during this period. The soft X-ray flux of the source went up and down three times on a timescale of about one month, as discovered by the ASM aboard RXTE, before finally returning to the normal level (of the hard state). The observed spectral and temporal X-ray properties of Cyg X-1 are mostly intermediate between the canonical hard and soft states. This is known previously for strong X-ray flares, however, we show that the source did enter a period that resembles, in many ways, a sustained soft state during the last of the three flares. We make detailed comparisons between this flare and the 1996 state transition, in terms of the observed X-ray properties, such as flux--hardness correlation, X-ray spectrum, and power density spectrum. We point out the similarities and differences, and discuss possible implications of the results on our understanding of the phenomena of flares and state transitions associated with Cyg X-1.

astro-ph↗