Unified Implicit Neural Stylization

Fan, Zhiwen; Jiang, Yifan; Wang, Peihao; Gong, Xinyu; Xu, Dejia; Wang, Zhangyang

doi:10.1007/978-3-031-19784-0_37

Cited by 32 publications

(3 citation statements)

References 63 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Downblocks reduce patch size and increase the channel length, whereas, convolution block (CB) follows instant normalization and leaky ReLU. 23 Downblock pass information from front to end to eradicate overfitting,. The upblock reconstructs the location information by joining corresponding downblock outputs.…”

Section: Network Architecturementioning

confidence: 99%

Estimation of uncertainty in Brain Tumor segmentation using modified multistage 3D-UNet on multimodal MRI images

Parmar,

Parikh

2024

JIST

View full text Add to dashboard Cite

Automated brain tumor segmentation is challenging due to the tumor tissues' shape, size, and appearance. Various methods used multi-mode MRI scans to segment sub-regions of brain tumors. 3D CNN methods improved performance in recent years, but most methods do not use uncertainty information in segmentation. For reliability and understanding, model prediction is vital for clinical decisions. This work studies three models namely 3D-UNet, Modified 3D-UNet, and Modified Multistage-3D-UNet for brain tumor segmentation. MRI volume bias correction and normalization were carried out using z-score normalization. Two patch generation strategies reduce memory use and class imbalance. Voxel-wise uncertainty evaluation was made for aleatoric and epistemic uncertainties using test time augmentation and dropout, respectively. Variance and entropy are used to measure the uncertainty of the modified multistage-3D-Unet segmentation model from ground truth. Variance creates separate uncertainty maps for each tumor sub-regions, whereas, entropy provides only global information. Uncertainty is used to filter miss-segmented predictions and improve accuracy. Uncertainty awareness increases model accuracy with dice scores of 0.93, 0.91, and 0.83 for tumor sub-regions WT, TC, and ET respectively.

show abstract

Section: Network Architecturementioning

confidence: 99%

Estimation of uncertainty in Brain Tumor segmentation using modified multistage 3D-UNet on multimodal MRI images

Parmar,

Parikh

2024

JIST

View full text Add to dashboard Cite

show abstract

“…Semantic-driven 3D scene editing is a much harder task compared with 2D photo editing because of the high demand for multi-view consistency, the scarcity of paired 3D data and its entangled geometry and appearance. Previous approaches either rely on laborious annotation (Kania et al 2022;Yang et al 2022), only support object deformation or translation (Tschernezki et al 2022;Kobayashi, Matsumoto, and Sitzmann 2022), or only perform global style transfer (Chen et al 2022;Chiang et al 2022;Fan et al 2022;Huang et al 2022) without strong semantic meaning. Recently, thanks to the development of score distillation sampling technique, text-guided editing has emerged as a promising direction with great potential.…”

Section: Related Workmentioning

confidence: 99%

FocalDreamer: Text-Driven 3D Editing via Focal-Fusion Assembly

Li,

Dou,

Shi

et al. 2024

AAAI

View full text Add to dashboard Cite

While text-3D editing has made significant strides in leveraging score distillation sampling, emerging approaches still fall short in delivering separable, precise and consistent outcomes that are vital to content creation. In response, we introduce FocalDreamer, a framework that merges base shape with editable parts according to text prompts for fine-grained editing within desired regions. Specifically, equipped with geometry union and dual-path rendering, FocalDreamer assembles independent 3D parts into a complete object, tailored for convenient instance reuse and part-wise control. We propose geometric focal loss and style consistency regularization, which encourage focal fusion and congruent overall appearance. Furthermore, FocalDreamer generates high-fidelity geometry and PBR textures which are compatible with widely-used graphics engines. Extensive experiments have highlighted the superior editing capabilities of FocalDreamer in both quantitative and qualitative evaluations.

show abstract

“…In this material, we provide a detailed description. Different from previous methods that derive accurate and expensive ground-truth depth maps by pretrained NeRF trained with dense views [12] or captured by high-accuracy depth scanners [11], our goal is to develop a good NeRF with cheap or even free coarse depth maps, i.e., pre-trained single-view depth estimation models and consumer-level depth sensors. To this end, we use Microsoft Azure Kinect, ZED 2, and iPhone 13 Pro to collect a new dataset NVS-RGBD.…”

Section: B Detailed Descriptions Of the New Dataset Nvs-rgbdmentioning

confidence: 99%

SparseNeRF: Distilling Depth Ranking for Few-shot Novel View Synthesis

Wang¹,

Chen²,

Loy³

et al. 2023

Preprint

View full text Add to dashboard Cite

MonoSDF* RegNeRF DSNeRF SparseNeRF (ours) Novel view (GT) Figure 1: Comparing the state-of-the-arts RegNeRF [29], MonoSDF* [49], DSNeRF [11], and our SparseNeRF with three views for training (* denotes a re-implementation for our task). RegNeRF regularizes geometry with sparsity and continuity constraints. MonoSDF, and DSNeRF use scale-invariant depth constraints supervised by coarse depth maps. Our SparseNeRF uses robust depth ranking of coarse depth maps. SparseNeRF can synthesize realistic novel views and coherent geometric depth (Please refer to the supplementary material for rendered videos).

show abstract

Unified Implicit Neural Stylization

Cited by 32 publications

References 63 publications

Estimation of uncertainty in Brain Tumor segmentation using modified multistage 3D-UNet on multimodal MRI images

Estimation of uncertainty in Brain Tumor segmentation using modified multistage 3D-UNet on multimodal MRI images

FocalDreamer: Text-Driven 3D Editing via Focal-Fusion Assembly

SparseNeRF: Distilling Depth Ranking for Few-shot Novel View Synthesis

Contact Info

Product

Resources

About