SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2503.14029 · CVPR · 2025

Rethinking End-to-End 2D to 3D Scene Segmentation in Gaussian Splatting

Qianyi Wu, Zhengzhe Liu, Chi-Wing Fu, Pheng-Ann Heng, Shi Qiu, Runsong Zhu, Ka-Hei Hui

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 11 functions out of this paper's own repositories and ran 5 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.

RepositoryRoleRan
runsong123/unified-lift canonical 5 of 11
FunctionStatusWhere it lives
boundary_iou Ran runsong123/unified-lift/script/eval_lerf_mask_unified_lift.py
code served (permissive licence) · get_code("c20005191e0618bd")
get_confience_map Ran runsong123/unified-lift/train_unified_lift.py
code served (permissive licence) · get_code("5f91cd17bff041c5")
l1_loss Ran runsong123/unified-lift/utils/loss_utils.py
code served (permissive licence) · get_code("ac0e42d6fbcfbbe6")
load_mask Ran runsong123/unified-lift/script/eval_lerf_mask_unified_lift.py
code served (permissive licence) · get_code("bf2d28223bdf332b")
mask_to_boundary Ran runsong123/unified-lift/script/eval_lerf_mask_unified_lift.py
code served (permissive licence) · get_code("b79a90dcf06de145")
feature_to_rgb Not yet run runsong123/unified-lift/render.py
code served (permissive licence) · get_code("955c607d3685039a")
id2rgb Not yet run runsong123/unified-lift/train_unified_lift.py
code served (permissive licence) · get_code("9acdd81f4e30bd46")
masked_l1_loss Not yet run runsong123/unified-lift/utils/loss_utils.py
code served (permissive licence) · get_code("0ac7968bd7da45d3")
show_mask Not yet run runsong123/unified-lift/ext/grounded_sam.py
code served (permissive licence) · get_code("eb026ce356b1b155")
visualize_obj Not yet run runsong123/unified-lift/train_unified_lift.py
code served (permissive licence) · get_code("74195777dbe870e4")
weighted_l1_loss Not yet run runsong123/unified-lift/utils/loss_utils.py
code served (permissive licence) · get_code("e14045a7568f1af5")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Lifting multi-view 2D instance segmentation to a radiance field has proven effective to enhance 3D understanding. Existing works rely on direct matching for end-to-end lifting, yielding inferior results, or employ a two-stage solution constrained by complex pre-or post-processing. In this work, we design Unified-Lift, a new end-to-end objectaware lifting approach that aims for high-quality 3D segmentation based on our object-aware 3D Gaussian representation. To start, we augment each Gaussian point with a Gaussian-level feature learned using a contrastive loss to encode instance information. Importantly, we introduce a learnable object-level codebook to account for individual objects in the scene for an explicit object-level understanding and associate the encoded object-level features with the Gaussian-level point features for segmentation predictions. While promising, achieving effective codebook learning is nontrivial and a naive solution leads to degraded performance. Hence, we formulate the association learning module and the noisy label filtering module for effective and robust codebook learning. We conduct experiments on three benchmarks LERF-Masked, Replica, and Messy Rooms. Both qualitative and quantitative results manifest that our Unified-Lift clearly outperforms existing methods in terms of segmentation quality and time efficiency.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2503.14029")
get_code_for_paper("2503.14029")
have("2503.14029")

Connect an agent — have() is free.