{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/understanding-convolution-for-semantic","title":"Understanding Convolution for Semantic Segmentation","arxiv_id":"1702.08502","date":"2017-02-27","proceeding":null,"authors":["Panqu Wang","Pengfei Chen","Ye Yuan","Ding Liu","Zehua Huang","Xiaodi Hou","Garrison Cottrell"],"abstract":"Recent advances in deep learning, especially deep convolutional neural\nnetworks (CNNs), have led to significant improvement over previous semantic\nsegmentation systems. Here we show how to improve pixel-wise semantic\nsegmentation by manipulating convolution-related operations that are of both\ntheoretical and practical value. First, we design dense upsampling convolution\n(DUC) to generate pixel-level prediction, which is able to capture and decode\nmore detailed information that is generally missing in bilinear upsampling.\nSecond, we propose a hybrid dilated convolution (HDC) framework in the encoding\nphase. This framework 1) effectively enlarges the receptive fields (RF) of the\nnetwork to aggregate global information; 2) alleviates what we call the\n\"gridding issue\" caused by the standard dilated convolution operation. We\nevaluate our approaches thoroughly on the Cityscapes dataset, and achieve a\nstate-of-art result of 80.1% mIOU in the test set at the time of submission. We\nalso have achieved state-of-the-art overall on the KITTI road estimation\nbenchmark and the PASCAL VOC2012 segmentation task. Our source code can be\nfound at https://github.com/TuSimple/TuSimple-DUC .","url_abs":"http://arxiv.org/abs/1702.08502v3","url_pdf":"http://arxiv.org/pdf/1702.08502v3.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"understanding-convolution-for-semantic","repo_url":"https://github.com/TuSimple/TuSimple-DUC","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"mxnet","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"understanding-convolution-for-semantic","repo_url":"https://github.com/leemathew1998/GradientWeight","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"understanding-convolution-for-semantic","repo_url":"https://github.com/leemathew1998/RG","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"understanding-convolution-for-semantic","repo_url":"https://github.com/modelhub-ai/duc-semantic","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"mxnet","reach":{"status":"ok"}},{"paper_slug":"understanding-convolution-for-semantic","repo_url":"https://github.com/y-ouali/pytorch_segmentation","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"segmentation","task_name":"Segmentation"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"},{"task_slug":"thermal-image-segmentation","task_name":"Thermal Image Segmentation"}],"methods":[{"method_slug":"1x1-convolution","method_name":"1x1 Convolution"},{"method_slug":"average-pooling","method_name":"Average Pooling"},{"method_slug":"batch-normalization","method_name":"Batch Normalization"},{"method_slug":"bottleneck-residual-block","method_name":"Bottleneck Residual Block"},{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"dilated-convolution","method_name":"Dilated Convolution"},{"method_slug":"global-average-pooling","method_name":"Global Average Pooling"},{"method_slug":"kaiming-initialization","method_name":"Kaiming Initialization"},{"method_slug":"max-pooling","method_name":"Max Pooling"},{"method_slug":"relu","method_name":"ReLU"},{"method_slug":"residual-block","method_name":"Residual Block"},{"method_slug":"residual-connection","method_name":"Residual Connection"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/semantic-segmentation-on-cityscapes","task":"Semantic Segmentation","dataset":"Cityscapes test","model":"DUC-HDC (ResNet-101)","rank_in_archive_order":63,"of":105,"metrics":{"Mean IoU (class)":"77.6%"},"uses_additional_data":false},{"leaderboard":"/sota/semantic-segmentation-on-pascal-voc-2012","task":"Semantic Segmentation","dataset":"PASCAL VOC 2012 test","model":"TuSimple","rank_in_archive_order":20,"of":51,"metrics":{"Mean IoU":"83.1%"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1702.08502","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1702.08502"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/leemathew1998/GradientWeight","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/leemathew1998/RG","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/y-ouali/pytorch_segmentation","reach":{"status":"unanswered"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/TuSimple/TuSimple-DUC","reach":{"status":"ok","spdx":"Apache-2.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/modelhub-ai/duc-semantic","reach":{"status":"ok"}}],"summary":{"unverified":1},"by_repo_kind":{"official":{"samples":1,"ran":0,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"2bbe696c78e01a2b","entry":"assureSingleInstanceName","repo":"TuSimple/TuSimple-DUC","repo_kind":"official","path":"tusimple_duc/core/cityscapes_labels.py","file_url":"https://github.com/TuSimple/TuSimple-DUC/blob/HEAD/tusimple_duc/core/cityscapes_labels.py","link_basis":"plan_row","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"2bbe696c78e01a2b"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}