{"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/arxiv-2510-20994","title":"VESSA: Video-based objEct-centric Self-Supervised Adaptation for Visual Foundation Models","arxiv_id":"2510.20994","date":"2025-10-23","proceeding":"NeurIPS","authors":["Jesimon Barreto","Carlos Caetano","André Araujo","William Robson Schwartz"],"abstract":"Foundation models have advanced computer vision by enabling strong performance across diverse tasks through large-scale pretraining and supervised fine-tuning. However, they may underperform in domains with distribution shifts and scarce labels, where supervised fine-tuning may be infeasible. While continued self-supervised learning for model adaptation is common for generative language models, this strategy has not proven effective for vision-centric encoder models. To address this challenge, we introduce a novel formulation of self-supervised fine-tuning for vision foundation models, where the model is adapted to a new domain without requiring annotations, leveraging only short multi-view object-centric videos. Our method is referred to as VESSA: Video-based objEct-centric Self-Supervised Adaptation for visual foundation models. VESSA's training technique is based on a self-distillation paradigm, where it is critical to carefully tune prediction heads and deploy parameter-efficient adaptation techniques - otherwise, the model may quickly forget its pretrained knowledge and reach a degraded state. VESSA benefits significantly from multi-view object observations sourced from different frames in an object-centric video, efficiently learning robustness to varied capture conditions, without the need of annotations. Through comprehensive experiments with 3 vision foundation models on 2 datasets, VESSA demonstrates consistent improvements in downstream classification tasks, compared to the base models and previous adaptation methods. Code is publicly available at https://github.com/jesimonbarreto/VESSA.","url_abs":"https://arxiv.org/abs/2510.20994","url_pdf":"https://arxiv.org/pdf/2510.20994","source":{"archive":null,"snapshot":"2025-07-28","note":"not in the Papers with Code archive (frozen at the snapshot)","row_kind":"graph","title_abstract_authors_date":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)"},"code_links":[],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2510.20994","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2510.20994"}},"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":[{"mentioned_in_github":null,"is_official":null,"provenance":"deterministic:regex_extraction","mentioned_in_paper":null,"url":"https://github.com/jesimonbarreto/VESSA","reach":{"status":"ok"}}],"summary":{"unverified":18},"by_repo_kind":{"found_in_text":{"samples":18,"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":18,"samples":[{"code_sha256_prefix":"010bedd967dcd102","entry":"calculate_means","repo":"jesimonbarreto/VESSA","repo_kind":"found_in_text","path":"src/trainer_vessa.py","file_url":"https://github.com/jesimonbarreto/VESSA/blob/HEAD/src/trainer_vessa.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"010bedd967dcd102"}},{"code_sha256_prefix":"21ae1e5ae268d699","entry":"distorted_bounding_box_crop","repo":"jesimonbarreto/VESSA","repo_kind":"found_in_text","path":"src/CO3D_dataset.py","file_url":"https://github.com/jesimonbarreto/VESSA/blob/HEAD/src/CO3D_dataset.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"21ae1e5ae268d699"}},{"code_sha256_prefix":"d4251401c82a2c42","entry":"generate_conditional_freeze_layers","repo":"jesimonbarreto/VESSA","repo_kind":"found_in_text","path":"src/trainer_vessa.py","file_url":"https://github.com/jesimonbarreto/VESSA/blob/HEAD/src/trainer_vessa.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"d4251401c82a2c42"}},{"code_sha256_prefix":"5856650dc88f7bc8","entry":"get_all_checkpoint_numbers","repo":"jesimonbarreto/VESSA","repo_kind":"found_in_text","path":"src/knn_main.py","file_url":"https://github.com/jesimonbarreto/VESSA/blob/HEAD/src/knn_main.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"5856650dc88f7bc8"}},{"code_sha256_prefix":"e8c6ddcc02a4b831","entry":"get_highest_checkpoint","repo":"jesimonbarreto/VESSA","repo_kind":"found_in_text","path":"src/knn_main.py","file_url":"https://github.com/jesimonbarreto/VESSA/blob/HEAD/src/knn_main.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"e8c6ddcc02a4b831"}},{"code_sha256_prefix":"531afd8c9c7c1127","entry":"get_param_from_pytorch_style_dict","repo":"jesimonbarreto/VESSA","repo_kind":"found_in_text","path":"src/load_params.py","file_url":"https://github.com/jesimonbarreto/VESSA/blob/HEAD/src/load_params.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"531afd8c9c7c1127"}},{"code_sha256_prefix":"2987a7b8e4efb760","entry":"load_tfrecords","repo":"jesimonbarreto/VESSA","repo_kind":"found_in_text","path":"src/datasets.py","file_url":"https://github.com/jesimonbarreto/VESSA/blob/HEAD/src/datasets.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"2987a7b8e4efb760"}},{"code_sha256_prefix":"60e06f935699f0bd","entry":"norm_kernel_init_fn","repo":"jesimonbarreto/VESSA","repo_kind":"found_in_text","path":"src/vit_vessa.py","file_url":"https://github.com/jesimonbarreto/VESSA/blob/HEAD/src/vit_vessa.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"60e06f935699f0bd"}},{"code_sha256_prefix":"14d0005d1bb5c05b","entry":"normalize","repo":"jesimonbarreto/VESSA","repo_kind":"found_in_text","path":"src/module_knn.py","file_url":"https://github.com/jesimonbarreto/VESSA/blob/HEAD/src/module_knn.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"14d0005d1bb5c05b"}},{"code_sha256_prefix":"9e3e4cfaccce750d","entry":"normalize_image","repo":"jesimonbarreto/VESSA","repo_kind":"found_in_text","path":"src/CO3D_dataset.py","file_url":"https://github.com/jesimonbarreto/VESSA/blob/HEAD/src/CO3D_dataset.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"9e3e4cfaccce750d"}},{"code_sha256_prefix":"05bce84f7c0629de","entry":"plot_example","repo":"jesimonbarreto/VESSA","repo_kind":"found_in_text","path":"src/trainer_vessa.py","file_url":"https://github.com/jesimonbarreto/VESSA/blob/HEAD/src/trainer_vessa.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"05bce84f7c0629de"}},{"code_sha256_prefix":"1eb29b62c712b5fa","entry":"preprocess_example","repo":"jesimonbarreto/VESSA","repo_kind":"found_in_text","path":"src/datasets.py","file_url":"https://github.com/jesimonbarreto/VESSA/blob/HEAD/src/datasets.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"1eb29b62c712b5fa"}},{"code_sha256_prefix":"107c2ad8b18f821d","entry":"preprocess_example_eval","repo":"jesimonbarreto/VESSA","repo_kind":"found_in_text","path":"src/datasets.py","file_url":"https://github.com/jesimonbarreto/VESSA/blob/HEAD/src/datasets.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"107c2ad8b18f821d"}},{"code_sha256_prefix":"30ba62bc00533432","entry":"preprocess_for_train","repo":"jesimonbarreto/VESSA","repo_kind":"found_in_text","path":"src/CO3D_dataset.py","file_url":"https://github.com/jesimonbarreto/VESSA/blob/HEAD/src/CO3D_dataset.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"30ba62bc00533432"}},{"code_sha256_prefix":"7f925d1fd1ac1751","entry":"restore_checkpoint","repo":"jesimonbarreto/VESSA","repo_kind":"found_in_text","path":"src/utils_vessa.py","file_url":"https://github.com/jesimonbarreto/VESSA/blob/HEAD/src/utils_vessa.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"7f925d1fd1ac1751"}},{"code_sha256_prefix":"b921833c63cd0f72","entry":"split_qkv_b","repo":"jesimonbarreto/VESSA","repo_kind":"found_in_text","path":"src/load_params.py","file_url":"https://github.com/jesimonbarreto/VESSA/blob/HEAD/src/load_params.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"b921833c63cd0f72"}},{"code_sha256_prefix":"0a6c9c38127ec83c","entry":"split_qkv_w","repo":"jesimonbarreto/VESSA","repo_kind":"found_in_text","path":"src/load_params.py","file_url":"https://github.com/jesimonbarreto/VESSA/blob/HEAD/src/load_params.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"0a6c9c38127ec83c"}},{"code_sha256_prefix":"a8c62c8d56e000e2","entry":"token_indexes_not_to_drop","repo":"jesimonbarreto/VESSA","repo_kind":"found_in_text","path":"src/vit_vessa.py","file_url":"https://github.com/jesimonbarreto/VESSA/blob/HEAD/src/vit_vessa.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"a8c62c8d56e000e2"}}]},"arxiv_metadata":{"licence":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)","fields":["title","abstract","authors","date"],"primary_category":"cs.CV","source":"arxiv_api"},"syntology_extracted_results":null}