C.J. Wu

Splannequin: Freezing Monocular Mannequin-Challenge Footage with Dual-Detection Splatting

Yiyao Huang, C.J. Wu, Wei‐Lun Chao, Yu-Lun Liu

Synthesizing high-fidelity frozen 3D scenes from monocular Mannequin-Challenge (MC) videos is a unique problem distinct from standard dynamic scene reconstruction. Instead of focusing on modeling motion, our goal is to create a frozen scene while strategically preserving subtle dynamics to enable user-controlled …

tw, us (code pays fourni par la source)

Skyfall-GS: Synthesizing Immersive 3D Urban Scenes from Satellite Imagery

Jie-Ying Lee, Yi Liu, Wei‐Cheng Chang, C.J. Wu et autres

tw, es, us (code pays fourni par la source)

FrugalNeRF: Fast Convergence for Extreme Few-shot Novel View Synthesis without Learned Priors

Chin-Yang Lin, C.J. Wu, Chang-Han Yeh, S.C.M. Yen et autres

Neural Radiance Fields (NeRF) face significant challenges in extreme few-shot scenarios, primarily due to overfitting and long training times. Existing methods, such as FreeNeRF and SparseNeRF, use frequency regularization or pre-trained priors but struggle with complex scheduling and bias. We introduce FrugalNeRF, …

tw, gb (code pays fourni par la source)

AuraFusion360: Augmented Unseen Region Alignment for Reference-based 360° Unbounded Scene Inpainting

C.J. Wu, Y. S. Chen, Jie-Ying Lee, Bo-Hsu Ke et autres

Three-dimensional scene inpainting is crucial for applications from virtual reality to architectural visualization, yet existing methods struggle with view consistency and geometric accuracy in 360° unbounded scenes. We present AuraFusion360, a novel reference-based method that enables high-quality object removal and hole filling …

tw (code pays fourni par la source)

AuraFusion360: Augmented Unseen Region Alignment for Reference-based 360° Unbounded Scene Inpainting

C.J. Wu, Y. S. Chen, Jie-Ying Lee, Bo-Hsu Ke et autres

Three-dimensional scene inpainting is crucial for applications from virtual reality to architectural visualization, yet existing methods struggle with view consistency and geometric accuracy in 360° unbounded scenes. We present AuraFusion360, a novel reference-based method that enables high-quality object removal and hole filling …

FrugalNeRF: Fast Convergence for Extreme Few-shot Novel View Synthesis without Learned Priors

Chin-Yang Lin, C.J. Wu, Chang-Han Yeh, S.C.M. Yen et autres

Neural Radiance Fields (NeRF) face significant challenges in extreme few-shot scenarios, primarily due to overfitting and long training times. Existing methods, such as FreeNeRF and SparseNeRF, use frequency regularization or pre-trained priors but struggle with complex scheduling and bias. We introduce FrugalNeRF, …

Two‐stage deep‐learning‐based colonoscopy polyp detection incorporating fisheye and reflection correction

Chen‐Ming Hsu, Tsung‐Hsing Chen, Chien‐Chang Hsu, C.J. Wu et autres

BACKGROUND AND AIM: Colonoscopy is a useful method for the diagnosis and management of colorectal diseases. Many computer-aided systems have been developed to assist clinicians in detecting colorectal lesions by analyzing colonoscopy images. However, fisheye-lens distortion and light reflection in colonoscopy images …

tw (code pays fourni par la source)

Performance comparison of fuzzy and proportional controllers in the application of image compression

C.J. Wu

In this work, the performance comparison of fuzzy controllers using the MOI (mean-of-inversion) defuzzification method and conventional proportional controllers with respect to accuracy and convergence speed in the application of image compression (JPEG) is given. The simulation results based on 15 USC …

tw (code pays fourni par la source)

Performance comparison of MOI and COA in the application of image library indices

C.J. Wu, Andrew H. Sung

In this work, the performance of the MOI (mean-of-inversion) method with respect to accuracy and convergence speed is compared with that of the COA (center-of-area) method in the application of image library indices in which only very rough images are needed and …

tw, us (code pays fourni par la source)

A conceptual associative memory with full memory capacity and ability to associate arbitrary patterns

C.J. Wu

In this paper, a conceptual associative memory (CAM), which uses the conceptual layer(s) to associate input/output layers, is proposed to achieve the most important requirement for neural networks as memory devices, that is, the ability to store arbitrary patterns in networks and …

us (code pays fourni par la source)

DART1-a possible ideal building block of memory systems

C.J. Wu

In this paper, a bidirectional associative memory dual adaptive resonance theory 1 (DART1) consisting of two simplified ART1 (SART1) networks, which share the same category layer called the conceptual layer and use only top-down weight matrices, is studied. The DART1 can meet …

us (code pays fourni par la source)

Performance comparison of the MOI, the COA, and the MOM methods in the application of JPEG

C.J. Wu, Andrew H. Sung

In this work, the performance of the MOI (mean-of inversion) method with respect to accuracy and convergence speed is compared with that of the COA (center-of-area) and the MOM (mean-of-maximum) methods in the application of JPEG (joint photographic experts group). The feature …

us (code pays fourni par la source)