SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2004.12989 · 2020

CoReNet: Coherent 3D scene reconstruction from a single RGB image

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 13 functions out of this paper's own repositories and ran 6 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
google-research/corenet canonical 6 of 13
FunctionStatusWhere it lives
dynamic_tile Ran google-research/corenet/src/corenet/misc_util.py
code served (permissive licence) · get_code("a5d01e7c869d2709")
flag Ran google-research/corenet/src/corenet/cmd_line_flags.py
code served (permissive licence) · get_code("804c188255bfaba2")
get_node_name Ran google-research/corenet/src/corenet/distributed.py
code served (permissive licence) · get_code("8c9fd1a495470a13")
parse_template_mapping Ran google-research/corenet/src/corenet/configuration.py
code served (permissive licence) · get_code("c6eb2e8508f55530")
safe_div Ran google-research/corenet/src/corenet/misc_util.py
code served (permissive licence) · get_code("b2b477cdb750bc2e")
to_tensor Ran google-research/corenet/src/corenet/misc_util.py
code served (permissive licence) · get_code("6a696c208af7bf1b")
gather Not yet run google-research/corenet/src/corenet/distributed.py
code served (permissive licence) · get_code("6d011553084806dd")
get_worker_range Not yet run google-research/corenet/src/corenet/distributed.py
code served (permissive licence) · get_code("85bd6f4985225c82")
is_gs_path Not yet run google-research/corenet/src/corenet/file_system.py
code served (permissive licence) · get_code("dee07c87d05f3ba5")
parse_flags Not yet run google-research/corenet/src/corenet/cmd_line_flags.py
code served (permissive licence) · get_code("d1f8b069d9e8a4ca")
parse_gs_path Not yet run google-research/corenet/src/corenet/file_system.py
code served (permissive licence) · get_code("da6018191f25fd4d")
print_tensor Not yet run google-research/corenet/src/corenet/debug_helpers.py
code served (permissive licence) · get_code("745563abe59e906b")
splitall Not yet run google-research/corenet/src/corenet/file_system.py
code served (permissive licence) · get_code("8b307f8cd05e69cc")

Repositories linked to this paper

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

Abstract

Advances in deep learning techniques have allowed recent work to reconstruct the shape of a single object given only one RBG image as input. Building on common encoder-decoder architectures for this task, we propose three extensions: (1) ray-traced skip connections that propagate local 2D information to the output 3D volume in a physically correct manner; (2) a hybrid 3D volume representation that enables building translation equivariant models, while at the same time encoding fine object details without an excessive memory footprint; (3) a reconstruction loss tailored to capture overall object geometry. Furthermore, we adapt our model to address the harder task of reconstructing multiple objects from a single image. We reconstruct all objects jointly in one pass, producing a coherent reconstruction, where all objects live in a single consistent 3D coordinate frame relative to the camera and they do not intersect in 3D space. We also handle occlusions and resolve them by hallucinating the missing object parts in the 3D volume. We validate the impact of our contributions experimentally both on synthetic data from ShapeNet as well as real images from Pix3D. Our method improves over the state-of-the-art single-object methods on both datasets. Finally, we evaluate performance quantitatively on multiple object reconstruction with synthetic scenes assembled from ShapeNet objects.

For agents

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

get_harvested_code_for_paper("2004.12989")
get_code_for_paper("2004.12989")
have("2004.12989")

Connect an agent — have() is free.