1-100 of about 247 matches for site:arxiv.org high-quality
[2306.09109] NAVI: Category-Agnostic Image Collections with High-Quality 3D Shape and Pose Annotatio
https://arxiv.org/abs/2306.09109
2306.09109] NAVI: Category-Agnostic Image Collections with High-Quality 3D Shape and Pose Annotations Skip to main
https://arxiv.org/abs/1905.10711
1905.10711] DISN: Deep Implicit Surface Network for High-quality Single-view 3D Reconstruction Skip to main
https://arxiv.org/abs/2404.12385
2404.12385] MeshLRM: Large Reconstruction Model for High-Quality Meshes Happy Open Access Week from arXiv! YOU make open
https://arxiv.org/abs/2311.17261
2311.17261] SceneTex: High-Quality Texture Synthesis for Indoor Scenes via Diffusion Priors Skip to main content
[2306.09109] NAVI: Category-Agnostic Image Collections with High-Quality 3D Shape and Pose Annotatio
https://arxiv.org/abs//2306.09109
2306.09109] NAVI: Category-Agnostic Image Collections with High-Quality 3D Shape and Pose Annotations Skip to main
https://arxiv.org/abs/2311.17261
2311.17261] SceneTex: High-Quality Texture Synthesis for Indoor Scenes via Diffusion Priors Skip to main content
[2306.09109] NAVI: Category-Agnostic Image Collections with High-Quality 3D Shape and Pose Annotatio
https://arxiv.org/abs//2306.09109
2306.09109] NAVI: Category-Agnostic Image Collections with High-Quality 3D Shape and Pose Annotations Skip to main
https://arxiv.org/abs/2409.12957
2409.12957] 3DTopia-XL: Scaling High-quality 3D Asset Generation via Primitive Diffusion The Scheduled Database Maintenance 2025-09
https://arxiv.org/abs/2409.12957
2409.12957] 3DTopia-XL: Scaling High-quality 3D Asset Generation via Primitive Diffusion Skip to main content We gratefully
[2310.01406] HumanNorm: Learning Normal Diffusion Model for High-quality and Realistic 3D Human Gene
https://arxiv.org/abs/2310.01406
2310.01406] HumanNorm: Learning Normal Diffusion Model for High-quality and Realistic 3D Human Generation Skip to
https://arxiv.org/abs/2411.07126
2411.07126] Edify Image: High-Quality Image Generation with Pixel Space Laplacian Diffusion Models Happy Open Access Week from arXiv! YOU
https://arxiv.org/abs/2506.07643
We introduce ROBIN: an MLM instruction-tuned with densely annotated relationships capable of constructing high-quality dense scene graphs
https://arxiv.org/abs/2310.13772
latent texture. We thoroughly validate TexFusion and show that we can efficiently generate diverse, high quality and globally
https://arxiv.org/abs/2506.07643
We introduce ROBIN: an MLM instruction-tuned with densely annotated relationships capable of constructing high-quality dense scene graphs
https://arxiv.org/abs/2012.05116
flash, in low-light environments. Our goal is to produce a high-quality rendering of
https://arxiv.org/abs/2010.02502
Song and 2 other authors View PDF Abstract: Denoising diffusion probabilistic models (DDPMs) have achieved high quality image generation without
https://arxiv.org/abs/2012.05116
flash, in low-light environments. Our goal is to produce a high-quality rendering of
https://arxiv.org/abs/2410.06231
experimental) Abstract: We propose RelitLRM, a Large Reconstruction Model (LRM) for generating high-quality Gaussian splatting representations
https://arxiv.org/abs/2103.00762
that this representation can be reconstructed using only multi-view image supervision and generates high-quality rendering results. More
https://arxiv.org/abs/2412.15689
distillation and consistency distillation to achieve few-step video generation, maintaining both high quality and diversity
https://arxiv.org/abs/2311.10709
for diffusion, and multi-stage training that enable us to directly generate high quality and high
https://arxiv.org/abs/2405.10314
7 other authors View PDF HTML (experimental) Abstract: Advances in 3D reconstruction have enabled high-quality 3D capture, but
https://arxiv.org/abs/2412.12463
for sampling synthetic pattern analogies, enables the creation of a large, high-quality synthetic training dataset
https://arxiv.org/abs/2406.09401
problems to be addressed in the future. Furthermore, we use this high-quality dataset to
https://arxiv.org/abs/2312.09250
the surfaces of 3D shapes, with the goal of synthesizing high-quality textures. Our approach
https://arxiv.org/abs/2303.01469
limitation, we propose consistency models, a new family of models that generate high quality samples by directly
https://arxiv.org/abs/2412.15689
distillation and consistency distillation to achieve few-step video generation, maintaining both high quality and diversity
https://arxiv.org/abs/2412.12463
for sampling synthetic pattern analogies, enables the creation of a large, high-quality synthetic training dataset
https://arxiv.org/abs/2505.10566
information into 3D space. We design a data generation pipeline to ensure high-quality 3D guidance throughout
[2305.15347] A Tale of Two Features: Stable Diffusion Complements DINO for Zero-Shot Semantic Corres
https://arxiv.org/abs/2305.15347
to-image diffusion models have made significant advances in generating and editing high-quality images. As a
https://arxiv.org/abs/2406.09417
that calibrating the text conditioning of the source distribution can produce high-quality generation and
https://arxiv.org/abs/2407.03162
suite, achieving higher success rates and reduced task completion times. Moreover, the high-quality teleoperation demonstrations improve
https://arxiv.org/abs/2305.15399
the internal patch distribution from a single 3D textured shape and generates high-quality variations with fine
https://arxiv.org/abs/2311.17061
time. In this paper, we propose an efficient yet effective framework, HumanGaussian, that generates high-quality 3D humans with
[2111.11215] Direct Voxel Grid Optimization: Super-fast Convergence for Radiance Fields Reconstructi
https://arxiv.org/abs/2111.11215
simple yet non-trivial techniques that contribute to fast convergence speed and high-quality output. First, we
https://arxiv.org/abs/2303.04803
Text-to-image diffusion models have the remarkable ability to generate high-quality images with diverse
https://arxiv.org/abs/2303.15951
is able to use the same perspective warping to render high-quality images on two
https://arxiv.org/abs/2503.16430
the strengths of both approaches, providing a promising direction for high-quality visual generation with
https://arxiv.org/abs/2306.07200
text-to-image synthesis models have achieved an exceptional level of photorealism, generating high-quality images from arbitrary
https://arxiv.org/abs/2312.02981
excel at rendering photorealistic novel views of complex scenes. However, recovering a high-quality NeRF typically requires
https://arxiv.org/abs/2406.09417
that calibrating the text conditioning of the source distribution can produce high-quality generation and
https://arxiv.org/abs/2303.04803
Text-to-image diffusion models have the remarkable ability to generate high-quality images with diverse
https://arxiv.org/abs/2303.15951
is able to use the same perspective warping to render high-quality images on two
https://arxiv.org/abs/2311.17061
time. In this paper, we propose an efficient yet effective framework, HumanGaussian, that generates high-quality 3D humans with
https://arxiv.org/abs/2104.08418
investigate the use of Neural Radiance Fields (NeRF) to learn high quality 3D object category
https://arxiv.org/abs/2404.19702
PDF Abstract: We propose GS-LRM, a scalable large reconstruction model that can predict high-quality 3D Gaussian primitives
https://arxiv.org/abs/2406.07520
a single image of any object and can synthesize an accurate, high-quality relit image under
[2502.09614] DexTrack: Towards Generalizable Neural Tracking Control for Dexterous Manipulation from
https://arxiv.org/abs/2502.09614
performance in dynamic environments. At the same time, to obtain high-quality tracking demonstrations, we
https://arxiv.org/abs/2406.09371
LRM-Zero, a Large Reconstruction Model (LRM) trained entirely on synthesized 3D data, achieving high-quality sparse-view 3D
https://arxiv.org/abs/2411.17249
normal training data. Instead of relying on large-scale annotated video datasets, we demonstrate high-quality video buffer estimation
https://arxiv.org/abs/2211.16677
train existing 2D diffusion models on these representations to generate 3D neural fields with high quality and diversity
[2301.07525] OmniObject3D: Large-Vocabulary 3D Object Dataset for Realistic Perception, Reconstructi
https://arxiv.org/abs/2301.07525
the real world, we propose OmniObject3D, a large vocabulary 3D object dataset with massive high-quality real-scanned 3D
https://arxiv.org/abs/2008.01815
360$^{\circ}$ images. MDPs are more compact than previous 3D scene representations and enable high-quality, efficient new view
https://arxiv.org/abs/2304.02744
ambiguous regions. Our optimization shares information between two poses, which allows us to produce high fidelity and realistic
https://arxiv.org/abs/2303.12074
have created a 3D GAN that is both efficient and of high quality, while allowing for
https://arxiv.org/abs/2403.17888
View PDF HTML (experimental) Abstract: 3D Gaussian Splatting (3DGS) has recently revolutionized radiance field reconstruction, achieving high quality novel view synthesis
https://arxiv.org/abs/2311.16854
1) 3D and 2D diffusion guidance to effectively learn a high-quality static 3D asset
https://arxiv.org/abs/2405.17414
authors View PDF HTML (experimental) Abstract: Research on video generation has recently made tremendous progress, enabling high-quality videos to
https://arxiv.org/abs/2208.01626
in the image. We present our results over diverse images and prompts, demonstrating high-quality synthesis and
https://arxiv.org/abs/2311.18828
Tianwei Yin and 6 other authors View PDF HTML (experimental) Abstract: Diffusion models generate high-quality images but require
https://arxiv.org/abs/2405.01796
in this work, we develop a completely automated process to generate high-quality topic pages for
https://arxiv.org/abs/2405.17414
authors View PDF HTML (experimental) Abstract: Research on video generation has recently made tremendous progress, enabling high-quality videos to
https://arxiv.org/abs/2412.09621
of obtaining ground truth annotations. We present a system for mining high-quality 4D reconstructions from
https://arxiv.org/abs/2406.00609
the problem of the shortage of large repositories of high-quality 3D training models
https://arxiv.org/abs/2003.12642
We introduce a novel learning-based method to reconstruct the high-quality geometry and
https://arxiv.org/abs/2008.03824
challenging effects like specularities, shadows and occlusions. This allows us to perform high-quality view synthesis and
https://arxiv.org/abs/2003.12642
We introduce a novel learning-based method to reconstruct the high-quality geometry and
[2305.15347] A Tale of Two Features: Stable Diffusion Complements DINO for Zero-Shot Semantic Corres
https://arxiv.org/abs/2305.15347
to-image diffusion models have made significant advances in generating and editing high-quality images. As a
https://arxiv.org/abs/2008.03824
challenging effects like specularities, shadows and occlusions. This allows us to perform high-quality view synthesis and
https://arxiv.org/abs/2311.06214
data. In this paper, we propose Instant3D, a novel method that generates high-quality and diverse
https://arxiv.org/abs/2312.11461
NeRF-based representations. However, a naive application of Gaussian splatting cannot generate high-quality animatable avatars and
https://arxiv.org/abs/2010.04595
that visual occlusions are implicitly taken into account. Extensive experiments demonstrate that our method can generate high-quality and realistic
https://arxiv.org/abs/2310.15110
the-shelf image diffusion models such as Stable Diffusion. Zero123++ excels in producing high-quality, consistent multi-view
https://arxiv.org/abs/2311.09217
object parts is required for generating diverse reconstructions with sharp textures. We also show high-quality text-to
https://arxiv.org/abs/2312.08885
We incorporate an RGBD panorama diffusion model to mitigate it, resulting in high-quality geometry. Extensive evaluation
https://arxiv.org/abs/2311.17857
scheme bypasses the need for view-inconsistent upsamplers and achieves high-quality multi-view consistent
https://arxiv.org/abs/2010.04595
that visual occlusions are implicitly taken into account. Extensive experiments demonstrate that our method can generate high-quality and realistic
https://arxiv.org/abs/2310.15110
the-shelf image diffusion models such as Stable Diffusion. Zero123++ excels in producing high-quality, consistent multi-view
https://arxiv.org/abs/2312.11461
NeRF-based representations. However, a naive application of Gaussian splatting cannot generate high-quality animatable avatars and
https://arxiv.org/abs/2104.08418
investigate the use of Neural Radiance Fields (NeRF) to learn high quality 3D object category
https://arxiv.org/abs/2312.02981
excel at rendering photorealistic novel views of complex scenes. However, recovering a high-quality NeRF typically requires
https://arxiv.org/abs/1809.09761
of 3D models but lack photorealistic appearance. We present an approach to automatically assign high-quality, realistic appearance models
https://arxiv.org/abs/2303.12074
have created a 3D GAN that is both efficient and of high quality, while allowing for
https://arxiv.org/abs/2311.09217
object parts is required for generating diverse reconstructions with sharp textures. We also show high-quality text-to
https://arxiv.org/abs/2405.01796
in this work, we develop a completely automated process to generate high-quality topic pages for
https://arxiv.org/abs/2311.18828
Tianwei Yin and 6 other authors View PDF HTML (experimental) Abstract: Diffusion models generate high-quality images but require
https://arxiv.org/abs/2212.03860
Somepalli and 4 other authors View PDF Abstract: Cutting-edge diffusion models produce images with high quality and customizability
https://arxiv.org/abs/2311.06214
data. In this paper, we propose Instant3D, a novel method that generates high-quality and diverse
https://arxiv.org/abs/2206.14797
synthesis and editing tasks. Recent advances in this field have also enabled high-quality 3D or video
https://arxiv.org/abs/2403.12409
Guidance, by Yongwei Chen and 5 other authors View PDF HTML (experimental) Abstract: Generating high-quality 3D assets from
https://arxiv.org/abs/2412.18565
advances in neural rendering, due to the scarcity of high-quality 3D datasets and
https://arxiv.org/abs/2112.07945
R. Chan and 10 other authors View PDF Abstract: Unsupervised generation of high-quality multi-view-consistent
https://arxiv.org/abs/2109.01349
the focus on dual-camera super-resolution (DCSR), which utilizes reference images for high-quality and high
https://arxiv.org/abs/2312.09168
and the initial diffusion noise map, which we utilize to consistently generate high-quality chrome balls. We
https://arxiv.org/abs/2406.07754
Comprehensive qualitative and quantitative evaluations demonstrate that HOI-Swap significantly outperforms existing methods, delivering high-quality video edits with
[2204.02232] IRON: Inverse Rendering by Optimizing Neural SDFs and Materials from Photometric Images
https://arxiv.org/abs/2204.02232
a neural inverse rendering pipeline called IRON that operates on photometric images and outputs high-quality 3D content in
https://arxiv.org/abs/2212.05032
on text-to-image synthesis (T2I) tasks. Despite their ability to generate high-quality yet creative images
https://arxiv.org/abs/2210.09276
our method on numerous inputs from various domains, showcasing a plethora of high quality complex semantic image