ICCV 2025 — corpus atlas
Caption labels, not panels. A six-panel figure counts once. Observed figures, tables and section spans describe the inspected source versions; they are not quotas or acceptance thresholds. For visual purpose, use the examples and type guides.
64 / 64inspected PDFs / oral inventory records
693physical pages in those PDFs
7editorial contribution types
0separate files, excluded from totals
What the measured manuscripts look like
Observed extraction results, not a target length or figure quota. The middle 50% is the interval from the 25th to the 75th percentile. Missing boundaries are excluded from that span’s denominator, not replaced by zero. Caption counts are extraction-assisted and can miss nonstandard or image-only labels; inspect each paper’s notes.
| Measure | Median | Middle 50% | Full range | Denominator |
|---|---|---|---|---|
| Inspected PDF lengthphysical pages | 11 | 10–11 | 9–13 | 64 papers |
| Before references / appendixestimated page-equivalents | 8.08 | 8.05–8.12 | 7.86–9.05 | 64 papers |
| Attached appendix spanestimated page-equivalents | Not resolved | Not resolved–Not resolved | Not resolved–Not resolved | 0 papers |
| Main figurescaption labels | 6 | 5–8 | 3–12 | 64 papers |
| Main tablescaption labels | 4 | 3–6 | 0–11 | 64 papers |
| Appendix figurescaption labels | 0 | 0–0 | 0–0 | 64 papers |
| Appendix tablescaption labels | 0 | 0–0 | 0–0 | 64 papers |
| Separate source-file lengthwhole-file physical pages, possibly including repeated main text; not appendix-only length | Not resolved | Not resolved–Not resolved | Not resolved–Not resolved | 0 papers |
| Separate source-file figureswhole-file caption labels; may include overlapping main-text displays | Not resolved | Not resolved–Not resolved | Not resolved–Not resolved | 0 papers |
| Separate source-file tableswhole-file caption labels; may include overlapping main-text displays | Not resolved | Not resolved–Not resolved | Not resolved–Not resolved | 0 papers |
Compare like contributions
Each measured paper has one primary editorial type so the totals sum to 64. Secondary contributions were not separately coded; empty secondary lists do not imply the absence of mixed contributions. These are classifications of the annotated argument, not official ICCV tracks. The table uses the same extraction method and shows medians; the category guides below explain what these artifacts should accomplish.
| Primary contribution | Papers | Main figures / tables | Body span ≈ pp. | Appendix span ≈ pp. |
|---|---|---|---|---|
| Architecture & representation | 23 | 6 / 4 | 8.1 | Not resolved0/23 attached boundaries; separate PDF median Not resolved pp. (0 papers) |
| Training, objectives & optimization | 9 | 6 / 6 | 8.06 | Not resolved0/9 attached boundaries; separate PDF median Not resolved pp. (0 papers) |
| Theory & formal analysis | 4 | 3.5 / 3.5 | 8.05 | Not resolved0/4 attached boundaries; separate PDF median Not resolved pp. (0 papers) |
| Empirical analysis, diagnosis & auditing | 5 | 8 / 5 | 8.09 | Not resolved0/5 attached boundaries; separate PDF median Not resolved pp. (0 papers) |
| Datasets, benchmarks & measurement | 4 | 5.5 / 6.5 | 8.14 | Not resolved0/4 attached boundaries; separate PDF median Not resolved pp. (0 papers) |
| Efficiency, inference & systems | 8 | 7.5 / 4.5 | 8.09 | Not resolved0/8 attached boundaries; separate PDF median Not resolved pp. (0 papers) |
| Domain-grounded methods & physical systems | 11 | 7 / 3 | 8.05 | Not resolved0/11 attached boundaries; separate PDF median Not resolved pp. (0 papers) |
Search all 64 measurement records
Open a paper’s ledger for every detected top-level section, its estimated length and share of the PDF, and figure/table locations. Its existing study explains what those visuals and sections do.
Showing 64 of 64 records.
| Paper / category | PDF pages | Main F / T | Appendix F / T | Body / appendix span ≈ pp. | Notes |
|---|---|---|---|---|---|
| GT-Loc: Unifying When and Where in Images Through a Joint Embedding SpaceTraining, objectives & optimization | 11 | 5 / 6 | 0 / 0 | 8.14 / Unresolved | No attached appendix boundary detected |
| Scaling Laws for Native Multimodal ModelsEmpirical analysis, diagnosis & auditing | 12 | 12 / 4 | 0 / 0 | 8.12 / Unresolved | No attached appendix boundary detected; Figure numbering discontinuity; see source noteSource correction retained |
| FixTalk: Taming Identity Leakage for High-Quality Talking Head Generation in Extreme CasesArchitecture & representation | 13 | 8 / 4 | 0 / 0 | 8.1 / Unresolved | No attached appendix boundary detected |
| Differentiable Room Acoustic Rendering with Multi-View Vision PriorsDomain-grounded methods & physical systems | 11 | 5 / 3 | 0 / 0 | 8.05 / Unresolved | No attached appendix boundary detectedSource correction retained |
| Token Activation Map to Visually Explain Multimodal LLMsArchitecture & representation | 11 | 8 / 4 | 0 / 0 | 8.05 / Unresolved | No attached appendix boundary detected |
| Multi-View 3D Point TrackingArchitecture & representation | 10 | 4 / 3 | 0 / 0 | 8.05 / Unresolved | No attached appendix boundary detected |
| Uncalibrated Structure from Motion on a SphereTheory & formal analysis | 10 | 4 / 3 | 0 / 0 | 8.04 / Unresolved | No attached appendix boundary detectedSource correction retained |
| Removing Cost Volumes from Optical Flow EstimatorsEfficiency, inference & systems | 11 | 8 / 5 | 0 / 0 | 8.13 / Unresolved | No attached appendix boundary detected |
| Image as an IMU: Estimating Camera Motion from a Single Motion-Blurred ImageDomain-grounded methods & physical systems | 10 | 7 / 3 | 0 / 0 | 8.05 / Unresolved | No attached appendix boundary detected |
| TrajectoryCrafter: Redirecting Camera Trajectory for Monocular Videos via Diffusion ModelsArchitecture & representation | 12 | 9 / 3 | 0 / 0 | 8.05 / Unresolved | No attached appendix boundary detected |
| RayZer: A Self-supervised Large View Synthesis ModelTraining, objectives & optimization | 12 | 6 / 7 | 0 / 0 | 8.08 / Unresolved | No attached appendix boundary detected |
| EVER: Exact Volumetric Ellipsoid Rendering for Real-time View SynthesisEfficiency, inference & systems | 10 | 7 / 3 | 0 / 0 | 8.05 / Unresolved | No attached appendix boundary detected |
| Self-Ensembling Gaussian Splatting for Few-Shot Novel View SynthesisTraining, objectives & optimization | 11 | 7 / 6 | 0 / 0 | 8.05 / Unresolved | No attached appendix boundary detected |
| Back on Track: Bundle Adjustment for Dynamic Scene ReconstructionArchitecture & representation | 10 | 7 / 4 | 0 / 0 | 8.11 / Unresolved | No attached appendix boundary detected |
| SceneSplat: Gaussian Splatting-based Scene Understanding with Vision-Language PretrainingDatasets, benchmarks & measurement | 12 | 6 / 9 | 0 / 0 | 9.05 / Unresolved | No attached appendix boundary detectedSource correction retained |
| Variance-Based Pruning for Accelerating and Compressing Trained NetworksEfficiency, inference & systems | 10 | 3 / 9 | 0 / 0 | 8.04 / Unresolved | No attached appendix boundary detected |
| Importance-Based Token Merging for Efficient Image and Video GenerationEfficiency, inference & systems | 13 | 8 / 9 | 0 / 0 | 8.09 / Unresolved | No attached appendix boundary detected |
| Knowledge Distillation for Learned Image CompressionEfficiency, inference & systems | 11 | 11 / 4 | 0 / 0 | 8.05 / Unresolved | No attached appendix boundary detected |
| Distilling Diffusion Models to Efficient 3D LiDAR Scene CompletionEfficiency, inference & systems | 10 | 5 / 4 | 0 / 0 | 8.1 / Unresolved | No attached appendix boundary detected |
| Heavy Labels Out! Dataset Distillation with Label Space LighteningEfficiency, inference & systems | 10 | 3 / 7 | 0 / 0 | 8.09 / Unresolved | No attached appendix boundary detected |
| RS-vHeat: Heat Conduction Guided Efficient Remote Sensing Foundation ModelArchitecture & representation | 12 | 5 / 9 | 0 / 0 | 8.1 / Unresolved | No attached appendix boundary detected |
| Towards a Unified Copernicus Foundation Model for Earth VisionDatasets, benchmarks & measurement | 12 | 5 / 5 | 0 / 0 | 8.16 / Unresolved | No attached appendix boundary detected |
| Learning Streaming Video Representation via Multitask TrainingTraining, objectives & optimization | 13 | 3 / 6 | 0 / 0 | 7.86 / Unresolved | No attached appendix boundary detected |
| LoftUp: Learning a Coordinate-Based Feature Upsampler for Vision Foundation ModelsArchitecture & representation | 11 | 8 / 6 | 0 / 0 | 8.1 / Unresolved | No attached appendix boundary detected |
| Learning Visual Hierarchies in Hyperbolic Space for Image RetrievalTraining, objectives & optimization | 11 | 6 / 4 | 0 / 0 | 8.06 / Unresolved | No attached appendix boundary detected |
| GMMamba: Group Masking Mamba for Whole Slide Image ClassificationArchitecture & representation | 10 | 5 / 8 | 0 / 0 | 8.05 / Unresolved | No attached appendix boundary detected |
| NullSwap: Proactive Identity Cloaking Against Deepfake Face SwappingDomain-grounded methods & physical systems | 10 | 3 / 5 | 0 / 0 | 8.1 / Unresolved | No attached appendix boundary detected |
| MaskControl: Spatio-Temporal Control for Masked Motion SynthesisArchitecture & representation | 11 | 4 / 5 | 0 / 0 | 8.05 / Unresolved | No attached appendix boundary detected |
| HairCUP: Hair Compositional Universal Prior for 3D Gaussian AvatarsArchitecture & representation | 11 | 8 / 1 | 0 / 0 | 8.1 / Unresolved | No attached appendix boundary detected |
| Understanding Co-speech Gestures in-the-wildDatasets, benchmarks & measurement | 11 | 4 / 7 | 0 / 0 | 8.13 / Unresolved | No attached appendix boundary detected |
| DPoser-X: Diffusion Model as Robust 3D Whole-body Human Pose PriorTraining, objectives & optimization | 10 | 8 / 11 | 0 / 0 | 8.15 / Unresolved | No attached appendix boundary detected |
| Teeth Reconstruction and Performance Capture Using a Phone CameraDomain-grounded methods & physical systems | 11 | 7 / 2 | 0 / 0 | 8.12 / Unresolved | No attached appendix boundary detected |
| Dynamic Typography: Bringing Text to Life via Video Diffusion PriorDomain-grounded methods & physical systems | 11 | 11 / 2 | 0 / 0 | 8.05 / Unresolved | No attached appendix boundary detected |
| Generating Physically Stable and Buildable Brick Structures from TextDomain-grounded methods & physical systems | 12 | 7 / 1 | 0 / 0 | 8.14 / Unresolved | No attached appendix boundary detected |
| WIR3D: Visually-Informed and Geometry-Aware 3D Shape AbstractionTraining, objectives & optimization | 12 | 11 / 2 | 0 / 0 | 8.04 / Unresolved | No attached appendix boundary detected |
| SparseFlex: High-Resolution and Arbitrary-Topology 3D Shape ModelingArchitecture & representation | 12 | 6 / 5 | 0 / 0 | 8.1 / Unresolved | No attached appendix boundary detected |
| ReCamMaster: Camera-Controlled Generative Rendering from A Single VideoArchitecture & representation | 11 | 6 / 4 | 0 / 0 | 8.15 / Unresolved | No attached appendix boundary detectedSource correction retained |
| Forecasting Continuous Non-Conservative Dynamical Systems in SO(3)Architecture & representation | 11 | 5 / 3 | 0 / 0 | 8.15 / Unresolved | No attached appendix boundary detected |
| Certifiably Optimal Anisotropic Rotation AveragingTheory & formal analysis | 10 | 3 / 1 | 0 / 0 | 8.05 / Unresolved | No attached appendix boundary detected |
| Deterministic Object Pose Confidence Region EstimationTheory & formal analysis | 10 | 5 / 4 | 0 / 0 | 8.05 / Unresolved | No attached appendix boundary detected |
| RePoseD: Efficient Relative Pose Estimation With Known Depth InformationTheory & formal analysis | 11 | 3 / 5 | 0 / 0 | 8.2 / Unresolved | No attached appendix boundary detected |
| Diving into the Fusion of Monocular Priors for Generalized Stereo MatchingArchitecture & representation | 11 | 4 / 7 | 0 / 0 | 8.05 / Unresolved | No attached appendix boundary detected |
| LaRender: Training-Free Occlusion Control in Image Generation via Latent RenderingArchitecture & representation | 10 | 6 / 3 | 0 / 0 | 8.05 / Unresolved | No attached appendix boundary detected |
| MikuDance: Animating Character Art with Mixed Motion DynamicsArchitecture & representation | 11 | 11 / 1 | 0 / 0 | 8.12 / Unresolved | No attached appendix boundary detected |
| Diffusion Transformer meets Multi-level Wavelet Spectrum for Single Image Super-ResolutionArchitecture & representation | 11 | 8 / 7 | 0 / 0 | 8.05 / Unresolved | No attached appendix boundary detected |
| LOTS of Fashion! Multi-Conditioning for Image Generation via Sketch-Text PairingArchitecture & representation | 10 | 4 / 2 | 0 / 0 | 8.21 / Unresolved | No attached appendix boundary detected |
| FlowEdit: Inversion-Free Text-Based Editing Using Pre-Trained Flow ModelsEmpirical analysis, diagnosis & auditing | 10 | 8 / 0 | 0 / 0 | 8.09 / Unresolved | No attached appendix boundary detected |
| LayerTracer: Cognitive-Aligned Layered SVG Synthesis via Diffusion TransformerDomain-grounded methods & physical systems | 11 | 9 / 3 | 0 / 0 | 8.05 / Unresolved | No attached appendix boundary detected |
| ROAR: Reducing Inversion Error in Generative Image WatermarkingDomain-grounded methods & physical systems | 10 | 7 / 5 | 0 / 0 | 8.05 / Unresolved | No attached appendix boundary detected |
| Moto: Latent Motion Token as the Bridging Language for Learning Robot Manipulation from VideosTraining, objectives & optimization | 12 | 12 / 3 | 0 / 0 | 8.09 / Unresolved | No attached appendix boundary detected |
| Automated Model Evaluation for Object Detection via Prediction Consistency and ReliabilityEmpirical analysis, diagnosis & auditing | 10 | 6 / 7 | 0 / 0 | 8.12 / Unresolved | No attached appendix boundary detectedSource correction retained |
| Counting Stacked ObjectsDomain-grounded methods & physical systems | 10 | 9 / 5 | 0 / 0 | 8.14 / Unresolved | No attached appendix boundary detectedSource correction retained |
| MIORe & VAR-MIORe: Benchmarks to Push the Boundaries of RestorationDatasets, benchmarks & measurement | 10 | 9 / 6 | 0 / 0 | 8.05 / Unresolved | No attached appendix boundary detected |
| Soft Local Completeness: Rethinking Completeness in XAITraining, objectives & optimization | 11 | 5 / 6 | 0 / 0 | 8.05 / Unresolved | No attached appendix boundary detected |
| SuperDec: 3D Scene Decomposition with Superquadrics PrimitivesArchitecture & representation | 11 | 10 / 3 | 0 / 0 | 8.04 / Unresolved | No attached appendix boundary detected |
| Diffusion Image PriorEmpirical analysis, diagnosis & auditing | 9 | 7 / 6 | 0 / 0 | 8.05 / Unresolved | No attached appendix boundary detected |
| Spatially-Varying AutofocusEfficiency, inference & systems | 10 | 12 / 1 | 0 / 0 | 8.08 / Unresolved | No attached appendix boundary detected |
| Towards Foundational Models for Single-Chip RadarEmpirical analysis, diagnosis & auditing | 11 | 9 / 5 | 0 / 0 | 8.05 / Unresolved | No attached appendix boundary detected |
| Event-based Visual VibrometryDomain-grounded methods & physical systems | 11 | 6 / 3 | 0 / 0 | 8.05 / Unresolved | No attached appendix boundary detected |
| CorrCLIP: Reconstructing Patch Correlations in CLIP for Open-Vocabulary Semantic SegmentationArchitecture & representation | 11 | 4 / 8 | 0 / 0 | 8.12 / Unresolved | No attached appendix boundary detected |
| E-SAM: Training-Free Segment Every Entity ModelDomain-grounded methods & physical systems | 10 | 7 / 5 | 0 / 0 | 8.05 / Unresolved | No attached appendix boundary detected |
| Online Reasoning Video Segmentation with Just-in-Time Digital TwinsArchitecture & representation | 9 | 4 / 6 | 0 / 0 | 8.05 / Unresolved | No attached appendix boundary detected |
| Easy3D: A Simple Yet Effective Method for 3D Interactive SegmentationArchitecture & representation | 10 | 5 / 4 | 0 / 0 | 8.1 / Unresolved | No attached appendix boundary detected |
| ForestFormer3D: A Unified Framework for End-to-End Segmentation of Forest LiDAR 3D Point CloudsArchitecture & representation | 11 | 4 / 3 | 0 / 0 | 8.11 / Unresolved | No attached appendix boundary detected |
No matching papers. Reset or use a shorter query.
Measurement method
The ledger measures the official main PDFs, including their reference pages, in adaptive two-column physical-page coordinates. It does not measure separate uninspected supplements. A missing attached-appendix boundary is not a zero-length supporting document.
Counts refer to detected figure and table labels, not panel counts, chart marks or scientific tasks. The same figure can perform several explanatory jobs, so the selected visual gallery is not a frequency taxonomy of every panel in the corpus.
Approximate section envelopes follow top-level heading boundaries and include floats and whitespace. They cannot establish exact word allocation or compliance with the official eight-page body rule. Source-bound corrections and audit notes remain separate from the authored writing interpretation.
Section spans run from one detected top-level heading to the next in physical-page coordinates. They include whitespace and floats, not exact text lengths, occupied-ink area or official page-limit accounting. Their percentages use the entire inspected PDF; the title, abstract and teaser can precede the first heading.
Missing boundaries remain unresolved rather than zero. Separately inspected files have separate denominators. Caption counts are extraction-assisted; contents entries, repeated labels, references and image-only captions need source checks. Corrections remain associated with the exact source URL.
Contribution labels are editorial, not official tracks. Selected examples do not measure visual-style prevalence, and accepted-paper observations cannot show which writing choices caused acceptance.
Download structural metadata · Inspect individual sources and studies