{"arxiv_id":"2407.12395","incoming":{"papers":3,"sentences":7,"groups":[{"kind":"contests","label":"Contests it","papers":1,"sentences":1,"quotes":[{"arxiv_id":"2503.20168","title":"EVolSplat: Efficient Volume-based Gaussian Splatting for Urban View Synthesis","url":"/paper/2503.20168","section":"Related Work","sentence":"PointNeRF [ref] and EDUS [ref] learn implicit radiance fields from dense input point clouds, limiting their practical applicability due to slow rendering.","intent_p":0.47,"contests_p":0.76}]},{"kind":"comparison","label":"Compares with it","papers":1,"sentences":2,"quotes":[{"arxiv_id":"2503.20168","title":"EVolSplat: Efficient Volume-based Gaussian Splatting for Urban View Synthesis","url":"/paper/2503.20168","section":"Experimental Setup","sentence":"Baselines We compare with several representative feed-forward methods, including MVSNeRF [ref] , MuRF [ref] , EDUS [ref] , MVSplat [ref] and PixelSplat [ref] .","intent_p":1.0,"contests_p":null}]},{"kind":"background","label":"Cites it as background","papers":2,"sentences":2,"quotes":[{"arxiv_id":"2504.11092","title":"Vivid4D: Improving 4D Reconstruction from Monocular Video by Video Inpainting","url":"/paper/2504.11092","section":"Related Work","sentence":"Recent advances in neural rendering, from NeRF [ref] to 3D-GS [ref] , have revolutionized static scene reconstruction and novel view synthesis.","intent_p":0.99,"contests_p":null},{"arxiv_id":"2412.01718","title":"HUGSIM: A Real-Time, Photo-Realistic and Closed-Loop Simulator for Autonomous Driving","url":"/paper/2412.01718","section":"Urban Scene Reconstruction","sentence":"These methods can be categorized into four classes: point-based [ref] , mesh-based [ref] , NeRF-based [ref] and 3DGS-based [ref] .","intent_p":1.0,"contests_p":null}]}],"unlabelled_sentences":2},"outgoing":{"papers":15,"sentences":26,"groups":[{"kind":"method_or_tool","label":"Uses the method or tool of","papers":6,"sentences":6,"quotes":[{"arxiv_id":"2303.16196","title":"SparseNeRF: Distilling Depth Ranking for Few-shot Novel View Synthesis","url":"/paper/2303.16196","section":"Experiment","sentence":"To maintain parity in our evaluation, we furnish the depth maps utilized in our method to train depth-supervision approaches [ref] , and employ our accumulated point cloud as the initialization for 3DGS [ref] .","intent_p":0.84,"contests_p":null},{"arxiv_id":"2106.10689","title":"NeuS: Learning Neural Implicit Surfaces by Volume Rendering for Multi-view Reconstruction","url":"/paper/2106.10689","section":"Implementation details","sentence":"Hierarchical Sampling For rendering a ray, we follow NeuS [ref] to perform hierarchical sampling during training and inference.","intent_p":1.0,"contests_p":null},{"arxiv_id":"2303.14167","title":"UrbanGIRAFFE: Representing Urban Scenes as Compositional Generative Neural Feature Fields","url":"/paper/2303.14167","section":"Depth Estimation:","sentence":"Instead, we draw inspiration from [ref] to employ a 3D Spatially-Adaptive Normalization convolutional neural network (SPADE CNN) to modulate feature volume generation at multiple scales, which is a popular strategy in semantic image synthesis.","intent_p":0.99,"contests_p":null},{"arxiv_id":"2306.04988","title":"StreetSurf: Extending Multi-view Implicit Surface Reconstruction to Street Views","url":"/paper/2306.04988","section":"Training and Fine-tuning","sentence":"Entropy regularization loss To penalize our model representing the distant landscape as semi-transparent, we introduce the entropy regularization loss to encourage opaque rendering followed by StreetSurf [ref] : [equation]","intent_p":0.95,"contests_p":null},{"arxiv_id":"2005.10821","title":"Hierarchical Multi-Scale Attention for Semantic Segmentation","url":"/paper/2005.10821","section":"Training and Fine-tuning","sentence":"Sky loss In order to separate the sky from the foreground and background solid structures, we leverage a pre-trained segmentation model [ref] to provide pseudo sky mask $M^sky$, and supervise the rendered sky mask with binary cross entropy (BCE) loss: [equation]","intent_p":1.0,"contests_p":null}]},{"kind":"comparison","label":"Compares with","papers":2,"sentences":5,"quotes":[{"arxiv_id":"2303.16196","title":"SparseNeRF: Distilling Depth Ranking for Few-shot Novel View Synthesis","url":"/paper/2303.16196","section":"Experiment","sentence":"Secondly, we compare with test-time optimization methods, including methods for sparse view settings: MixNeRF [ref] , SparseNeRF [ref] , DS-NeRF [ref] , and one taking point clouds as input: 3DGS [ref] .","intent_p":0.95,"contests_p":null},{"arxiv_id":"2312.04565","title":"MuRF: Multi-Baseline Radiance Fields","url":"/paper/2312.04565","section":"Experiment","sentence":"Firstly, we compare our model with generalizable approaches: IBRNet [ref] , MVSNeRF [ref] , NeO 360 [ref] and MuRF [ref] .","intent_p":1.0,"contests_p":null}]},{"kind":"supports_this_paper","label":"Cites as support","papers":1,"sentences":1,"quotes":[{"arxiv_id":"2312.04565","title":"MuRF: Multi-Baseline Radiance Fields","url":"/paper/2312.04565","section":"Related Work","sentence":"As suggested in the current work MuRF [ref] , these methods struggle to generalize well to large baselines.","intent_p":0.97,"contests_p":null}]},{"kind":"background","label":"Cites as background","papers":8,"sentences":11,"quotes":[{"arxiv_id":"2303.16196","title":"SparseNeRF: Distilling Depth Ranking for Few-shot Novel View Synthesis","url":"/paper/2303.16196","section":"Introduction","sentence":"Several recent methods address the sparse view setting: One line of works [ref] exploits external geometric priors or regularization terms, such as depth maps, geometric continuity, and color cues, to enhance model performance.","intent_p":0.83,"contests_p":null},{"arxiv_id":"2401.10891","title":"Depth Anything: Unleashing the Power of Large-Scale Unlabeled Data","url":"/paper/2401.10891","section":"Related Work","sentence":"[ref] learn a scene starting from a coarse point cloud, which can be obtained from multi-view stereo techniques [ref] or learning-based methods [ref] by unprojecting depth maps.","intent_p":0.83,"contests_p":null},{"arxiv_id":"2312.04565","title":"MuRF: Multi-Baseline Radiance Fields","url":"/paper/2312.04565","section":"Related Work","sentence":"Generalizable NeRF approaches are also intensively studied in the literature [ref] .","intent_p":0.81,"contests_p":null},{"arxiv_id":"2303.00749","title":"S-NeRF: Neural Radiance Fields for Street Views","url":"/paper/2303.00749","section":"Related Work","sentence":"Street Scene NeRF There are a few NeRF-based methods carefully designed to deal with street scenes [ref] .","intent_p":0.94,"contests_p":null},{"arxiv_id":"2311.02077","title":"EmerNeRF: Emergent Spatial-Temporal Scene Decomposition via Self-Supervision","url":"/paper/2311.02077","section":"Introduction","sentence":"Several works [ref] subsequently demonstrate promising NVS performance for street scenes.","intent_p":0.83,"contests_p":null}]}],"unlabelled_sentences":3},"claims":[],"thresholds":{"intent_p":0.8,"supports_p":0.95,"contests_p":0.7},"method":"Citing sentences come from each paper's LaTeX source; every \\cite marks the sentence and the works it cites. Pith's classifier typed each sentence. A label is shown at intent probability 0.8 or more (0.95 for support), a contest at 0.7 or more. Measured accuracy against two blind model families is at https://pith.science/accuracy. The quoted sentences are the citing authors' own.","held_up":null,"_links":{"html":"https://pith.science/paper/2407.12395"}}