REVIEW 22 cited by
NeRF--: Neural Radiance Fields Without Known Camera Parameters
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
abstract
Considering the problem of novel view synthesis (NVS) from only a set of 2D images, we simplify the training process of Neural Radiance Field (NeRF) on forward-facing scenes by removing the requirement of known or pre-computed camera parameters, including both intrinsics and 6DoF poses. To this end, we propose NeRF$--$, with three contributions: First, we show that the camera parameters can be jointly optimised as learnable parameters with NeRF training, through a photometric reconstruction; Second, to benchmark the camera parameter estimation and the quality of novel view renderings, we introduce a new dataset of path-traced synthetic scenes, termed as Blender Forward-Facing Dataset (BLEFF); Third, we conduct extensive analyses to understand the training behaviours under various camera motions, and show that in most scenarios, the joint optimisation pipeline can recover accurate camera parameters and achieve comparable novel view synthesis quality as those trained with COLMAP pre-computed camera parameters. Our code and data are available at https://nerfmm.active.vision.
Forward citations
Cited by 22 Pith papers
-
CalibAnyView: Beyond Single-View Camera Calibration in the Wild
A multi-view transformer predicts dense perspective fields that feed a geometric optimizer to estimate camera intrinsics and gravity from arbitrary numbers of real-world views.
-
HairGPT: Strand-as-Language Autoregressive Modeling for Realistic 3D Hairstyle Synthesis
HairGPT reframes 3D hairstyle synthesis as dual-decoupled autoregressive strand sequence modeling with geometric tokenization for semantic control and rare style generation.
-
NoDrift3R: Raymap-Guided Coupling for Drift-Robust Unposed Feed-Forward 3D Reconstruction
Anchoring Gaussian centers to predicted raymaps and jointly optimizing RGB, raymap, and camera losses with a dual-frequency curriculum suppresses pose drift and improves pose-free 3D reconstruction on long sequences.
-
StructSplat: Generalizable 3D Gaussian Splatting from Uncalibrated Sparse Views
StructSplat introduces a structured 3D Gaussian splatting framework that performs feed-forward reconstruction from uncalibrated sparse views using pixel-aligned features, semantic priors, and camera alignment.
-
RayFormer: Modeling Inter- and Intra-Ray Similarity for NeRF-Based Video Snapshot Compressive Imaging
RayFormer improves NeRF reconstruction for video SCI by replacing random ray sampling with patch-level sampling, adding a transformer to capture inter- and intra-ray structural similarities, and incorporating a total ...
-
PCM-NeRF: Probabilistic Camera Modeling for Neural Radiance Fields under Pose Uncertainty
PCM-NeRF improves neural surface reconstruction under uncertain camera poses by learning per-camera pose distributions and damping updates from high-uncertainty views.
-
LiveStre4m: Feed-Forward Live Streaming of Novel Views from Unposed Multi-View Video
LiveStre4m delivers real-time novel-view video streaming from unposed multi-view inputs via a multi-view vision transformer, diffusion-transformer interpolation, and a learned camera pose predictor.
-
LongSplat: Robust Unposed 3D Gaussian Splatting for Casual Long Videos
An incremental 3D Gaussian Splatting pipeline that jointly optimizes camera poses and scene geometry using MASt3R priors and density-adaptive octree anchors achieves state-of-the-art novel view synthesis on casual lon...
-
The Less You Depend, The More You Learn: Synthesizing Novel Views from Sparse, Unposed Images with Minimal 3D Knowledge
Data-centric novel view synthesis models with minimal 3D knowledge and no pose annotations scale better with data volume and outperform traditional bias-driven methods.
-
RoDyGS: Robust Dynamic Gaussian Splatting for Casual Videos
RoDyGS separates static and dynamic elements in monocular videos using Gaussian splatting with regularization and introduces the Kubric-MRig benchmark for pose-free dynamic novel view synthesis.
-
SalientGS: Unified SfM-to-3DGS with Importance-Guided MCMC Gaussian Allocation
Importance-guided MCMC reallocates 3D Gaussians toward multi-view underfit regions, enabling a unified SfM-to-3DGS pipeline that finishes in ~15 minutes with SOTA perceptual quality.
-
SalientGS: Unified SfM-to-3DGS with Importance-Guided MCMC Gaussian Allocation
SalientGS integrates fast first-order SfM, joint pose refinement, and importance-guided MCMC Gaussian birth/relocation to reach 27.65 dB macro-average PSNR at 1.5M Gaussians in about 10 minutes end-to-end.
-
NoDrift3R: Raymap-Guided Coupling for Drift-Robust Unposed Feed-Forward 3D Reconstruction
Anchoring 3D Gaussian centers to ray-map predictions and jointly optimizing geometry with appearance supervision suppresses pose drift in unposed feed-forward 3D reconstruction.
-
MZEN: Multi-Zoom Enhanced NeRF for 3-D Reconstruction with Unknown Camera Poses
MZEN is a NeRF training schedule that handles multi-zoom image sets with a zoom-scaled camera model and a bootstrap-register-refine pose strategy, claiming large gains on a new eight-scene benchmark.
-
KFC-W: Generating 3D-Consistent Videos from Unposed Internet Photos
KFC-W is a self-supervised 3D-aware video model trained on videos and multiview internet photos that produces geometrically consistent interpolations between unposed input images without any 3D annotations.
-
Splatt3R: Zero-shot Gaussian Splatting from Uncalibrated Image Pairs
Splatt3R is a feed-forward network that predicts 3D Gaussian splats directly from uncalibrated stereo image pairs by extending MASt3R with appearance attributes and a two-stage training procedure.
-
BSNeRF: Broadband Spectral Neural Radiance Fields for Snapshot Multispectral Light-field Imaging
A neural radiance field with joint camera pose estimation is proposed to decouple broadband spectral multiplexing in snapshot multispectral light-field images.
-
Embodied Spatial Intelligence: from Implicit Scene Modeling to Spatial Reasoning
The thesis demonstrates that combining implicit 3D scene representations with LLM-based reasoning, using text as an interface, yields strong performance on robotic perception and spatial language tasks.
-
DrivingGaussian++: Towards Realistic Reconstruction and Editable Simulation for Surrounding Dynamic Driving Scenes
DrivingGaussian++ reconstructs dynamic surround-view driving scenes and performs training-free multi-task editing (weather, texture, object manipulation) using Gaussians, diffusion models, and LLM-generated trajectories.
-
Novel View Synthesis with Gaussian Splatting: Impact on Photogrammetry Model Accuracy and Resolution
Gaussian Splatting beats photogrammetry on image quality metrics, and adding its synthetic views to a photogrammetry dataset raises SSIM/PSNR but lowers measured resolution.
-
Neural Field Representations of Mobile Computational Photography
Fitting neural fields directly to raw phone bursts reconstructs depth, separates reflections and occluders, and stitches panoramas, outperforming the compared baselines on the thesis's benchmarks.
-
NeRF: Neural Radiance Field in 3D Vision: A Comprehensive Review (Updated Post-Gaussian Splatting)
A literature survey of NeRF and neural field methods from 2020-2025, organized by architecture and application taxonomies with benchmarks and dataset overviews, covering both pre- and post-Gaussian Splatting periods.
Discussion (0). Sign in to comment.