{"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/high-performance-transformers-for-table","title":"High-Performance Transformers for Table Structure Recognition Need Early Convolutions","arxiv_id":"2311.05565","date":"2023-11-09","proceeding":null,"authors":["Shengyun Peng","Seongmin Lee","XiaoJing Wang","Rajarajeswari Balasubramaniyan","Duen Horng Chau"],"abstract":"Table structure recognition (TSR) aims to convert tabular images into a machine-readable format, where a visual encoder extracts image features and a textual decoder generates table-representing tokens. Existing approaches use classic convolutional neural network (CNN) backbones for the visual encoder and transformers for the textual decoder. However, this hybrid CNN-Transformer architecture introduces a complex visual encoder that accounts for nearly half of the total model parameters, markedly reduces both training and inference speed, and hinders the potential for self-supervised learning in TSR. In this work, we design a lightweight visual encoder for TSR without sacrificing expressive power. We discover that a convolutional stem can match classic CNN backbone performance, with a much simpler model. The convolutional stem strikes an optimal balance between two crucial factors for high-performance TSR: a higher receptive field (RF) ratio and a longer sequence length. This allows it to \"see\" an appropriate portion of the table and \"store\" the complex table structure within sufficient context length for the subsequent transformer. We conducted reproducible ablation studies and open-sourced our code at https://github.com/poloclub/tsr-convstem to enhance transparency, inspire innovations, and facilitate fair comparisons in our domain as tables are a promising modality for representation learning.","url_abs":"https://arxiv.org/abs/2311.05565v1","url_pdf":"https://arxiv.org/pdf/2311.05565v1.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":"high-performance-transformers-for-table","repo_url":"https://github.com/poloclub/tsr-convstem","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"high-performance-transformers-for-table","repo_url":"https://github.com/poloclub/unitable","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"decoder","task_name":"Decoder"},{"task_slug":"representation-learning","task_name":"Representation Learning"},{"task_slug":"self-supervised-learning","task_name":"Self-Supervised Learning"},{"task_slug":"table-recognition","task_name":"Table Recognition"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/table-recognition-on-pubtabnet","task":"Table Recognition","dataset":"PubTabNet","model":"ConvStem","rank_in_archive_order":4,"of":13,"metrics":{"TEDS (all samples)":"96.53"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2311.05565","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2311.05565"}},"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":"deterministic:regex_extraction","url":"https://github.com/poloclub/tsr-convstem","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/poloclub/unitable","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran":1,"unverified":1},"by_repo_kind":{"official":{"samples":2,"ran":1,"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":"de9e6c4743b42946","entry":"parallel_process","repo":"poloclub/tsr-convstem","repo_kind":"official","path":"src/utils/teds.py","file_url":"https://github.com/poloclub/tsr-convstem/blob/HEAD/src/utils/teds.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"de9e6c4743b42946"}},{"code_sha256_prefix":"eaa8420ba2c27ae6","entry":"combine_html_table_structure_text","repo":"poloclub/tsr-convstem","repo_kind":"official","path":"src/utils/utils.py","file_url":"https://github.com/poloclub/tsr-convstem/blob/HEAD/src/utils/utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"eaa8420ba2c27ae6"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}