{"corpus_id":208629777,"paper_sha":"bce087b969a1d6a7b6771c6cbf30a4f2cacaea64","doi":"10.1109/ITSC.2019.8916914","arxiv_id":null,"pmid":null,"pmcid":null,"mag_id":2990989819,"dblp_id":"conf/itsc/GuoASMO19","acl_id":null,"title":"Monocular Depth Estimation Using Synthetic Images With Shadow Removal*","year":2019,"publication_date":"2019-10-01","venue":"International Conference on Intelligent Transportation Systems","journal":{"name":"2019 IEEE Intelligent Transportation Systems Conference (ITSC)","pages":"1432-1439","volume":null},"journal_issn":null,"journal_title":null,"publication_types":["JournalArticle","Conference"],"pubmed_pub_types":null,"s2_fields_of_study":["Computer Science","Engineering"],"reference_count":50,"citation_count":16,"influential_citation_count":0,"is_open_access":false,"arxiv_categories":null,"arxiv_license":null,"arxiv_journal_ref":null,"mesh_headings":null,"chemicals":null,"comments_corrections":null,"source_flags":1,"s2_open_access_pdf_url":null,"s2_open_access_landing_url":null,"s2_open_access_license":null,"s2_open_access_status":null,"pmc_open_access_pdf_url":null,"pmc_open_access_landing_url":null,"pmc_open_access_license":null,"pmc_open_access_status":null,"unpaywall_open_access_pdf_url":null,"unpaywall_open_access_landing_url":null,"unpaywall_open_access_license":null,"unpaywall_open_access_status":null,"abstract":"Learning based monocular depth estimation has become popular in recent years. However, training of reliable depth estimators requires large volumes of ground truth depth data, which is expensive to obtain. To overcome this challenge, we propose a novel monocular depth estimation system that adopts pixel-perfect synthetic image training. The model is trained with synthetic data but inferred with realistic images by applying image domain adaptation. Considering realistic constraints, such as shadow regions, which cause the performance drops in depth estimation, the system adopts a dedicated module to remove such ambient shadows from images to guarantee premier performance in the task. Experimental results, with both synthetic and realistic benchmarks, indicate the efficacy and the advantages of the system compared to existing state-of-the-art technologies.","claims":[{"public_id":"cl_cbbe5ef56fa17e8786f852403b8e59e9","status":"active","text":"A dedicated shadow removal module removes ambient shadows from images to reduce performance drops in depth estimation under shadow regions.","confidence":0.9,"contributors":[{"id":136,"public_id":"3c2apqe3ut","public_label":"Anonymous (3c2apqe3ut)","roles":["extraction"],"url":"https://sah.borca.ai/u/3c2apqe3ut"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/claims/cl_cbbe5ef56fa17e8786f852403b8e59e9"},{"public_id":"cl_5f1e3ef1bd14b64c885e0f27a5444ded","status":"active","text":"A monocular depth estimation system uses pixel-perfect synthetic image training with image domain adaptation to train on synthetic data and infer on realistic images.","confidence":0.92,"contributors":[{"id":136,"public_id":"3c2apqe3ut","public_label":"Anonymous (3c2apqe3ut)","roles":["extraction"],"url":"https://sah.borca.ai/u/3c2apqe3ut"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/claims/cl_5f1e3ef1bd14b64c885e0f27a5444ded"},{"public_id":"cl_0d61811beefc537589710fbd5dfa587b","status":"active","text":"Experimental results on synthetic and realistic benchmarks indicate advantages over existing state-of-the-art technologies.","confidence":0.86,"contributors":[{"id":136,"public_id":"3c2apqe3ut","public_label":"Anonymous (3c2apqe3ut)","roles":["extraction"],"url":"https://sah.borca.ai/u/3c2apqe3ut"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/claims/cl_0d61811beefc537589710fbd5dfa587b"}],"concepts":[{"public_id":"co_07b07f8898c5ce2b209c83df2cf27d3a","status":"active","name":"shadow regions","description":"Image areas affected by shadows that create realistic constraints for depth estimation.","types":["phenomenon"],"aliases":[],"contributors":[{"id":136,"public_id":"3c2apqe3ut","public_label":"Anonymous (3c2apqe3ut)","roles":["extraction"],"url":"https://sah.borca.ai/u/3c2apqe3ut"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/concepts/co_07b07f8898c5ce2b209c83df2cf27d3a"},{"public_id":"co_6e38fcbdff12bf8fe114ed6e819403f1","status":"active","name":"state-of-the-art technologies","description":"Existing leading depth estimation technologies used as comparison baselines.","types":["baseline"],"aliases":[],"contributors":[{"id":136,"public_id":"3c2apqe3ut","public_label":"Anonymous (3c2apqe3ut)","roles":["extraction"],"url":"https://sah.borca.ai/u/3c2apqe3ut"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/concepts/co_6e38fcbdff12bf8fe114ed6e819403f1"},{"public_id":"co_73baa36b7a50170ea672d7069602d55c","status":"active","name":"synthetic data","description":"Artificially generated image data used to train the depth estimation model.","types":["data"],"aliases":[],"contributors":[{"id":136,"public_id":"3c2apqe3ut","public_label":"Anonymous (3c2apqe3ut)","roles":["extraction"],"url":"https://sah.borca.ai/u/3c2apqe3ut"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/concepts/co_73baa36b7a50170ea672d7069602d55c"},{"public_id":"co_75324c9f0f3922e6b80d4d91fe83b5e0","status":"active","name":"synthetic and realistic benchmarks","description":"Evaluation settings using both synthetic and realistic data to test the system.","types":["evaluation setting"],"aliases":[],"contributors":[{"id":136,"public_id":"3c2apqe3ut","public_label":"Anonymous (3c2apqe3ut)","roles":["extraction"],"url":"https://sah.borca.ai/u/3c2apqe3ut"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/concepts/co_75324c9f0f3922e6b80d4d91fe83b5e0"},{"public_id":"co_78df7b7ec7a486461410e14fffcbd78b","status":"active","name":"realistic images","description":"Images from realistic domains used for inference by the trained model.","types":["data"],"aliases":[],"contributors":[{"id":136,"public_id":"3c2apqe3ut","public_label":"Anonymous (3c2apqe3ut)","roles":["extraction"],"url":"https://sah.borca.ai/u/3c2apqe3ut"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/concepts/co_78df7b7ec7a486461410e14fffcbd78b"},{"public_id":"co_845a8c4512f458cc9cb6f004623e4dc2","status":"active","name":"ground truth depth data","description":"Reference depth measurements used as supervision for training depth estimators.","types":["data"],"aliases":[],"contributors":[{"id":136,"public_id":"3c2apqe3ut","public_label":"Anonymous (3c2apqe3ut)","roles":["extraction"],"url":"https://sah.borca.ai/u/3c2apqe3ut"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/concepts/co_845a8c4512f458cc9cb6f004623e4dc2"},{"public_id":"co_9bb451392200d25f9ec0982351aef01d","status":"active","name":"monocular depth estimation system","description":"The proposed system for estimating depth from one image using synthetic training and adaptation components.","types":["system"],"aliases":[],"contributors":[{"id":136,"public_id":"3c2apqe3ut","public_label":"Anonymous (3c2apqe3ut)","roles":["extraction"],"url":"https://sah.borca.ai/u/3c2apqe3ut"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/concepts/co_9bb451392200d25f9ec0982351aef01d"},{"public_id":"co_9d63057b149437775838fbeaa1487d7c","status":"active","name":"shadow removal module","description":"A dedicated component of the system that removes ambient shadows from images.","types":["module"],"aliases":[],"contributors":[{"id":136,"public_id":"3c2apqe3ut","public_label":"Anonymous (3c2apqe3ut)","roles":["extraction"],"url":"https://sah.borca.ai/u/3c2apqe3ut"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/concepts/co_9d63057b149437775838fbeaa1487d7c"},{"public_id":"co_a58289147dad75e4427abdd9f679ba83","status":"active","name":"ambient shadows","description":"Shadows present in images that the system targets for removal before depth estimation.","types":["phenomenon"],"aliases":[],"contributors":[{"id":136,"public_id":"3c2apqe3ut","public_label":"Anonymous (3c2apqe3ut)","roles":["extraction"],"url":"https://sah.borca.ai/u/3c2apqe3ut"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/concepts/co_a58289147dad75e4427abdd9f679ba83"},{"public_id":"co_aa7560fb279db6562a3aeeb682596ae6","status":"active","name":"image domain adaptation","description":"A method for adapting a model trained on synthetic images to operate on realistic images.","types":["method"],"aliases":[],"contributors":[{"id":136,"public_id":"3c2apqe3ut","public_label":"Anonymous (3c2apqe3ut)","roles":["extraction"],"url":"https://sah.borca.ai/u/3c2apqe3ut"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/concepts/co_aa7560fb279db6562a3aeeb682596ae6"},{"public_id":"co_bb9ad52aa7012efd28900b89f8356405","status":"active","name":"monocular depth estimation","description":"A computer vision task that estimates scene depth from a single image.","types":["task"],"aliases":[],"contributors":[{"id":136,"public_id":"3c2apqe3ut","public_label":"Anonymous (3c2apqe3ut)","roles":["extraction"],"url":"https://sah.borca.ai/u/3c2apqe3ut"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/concepts/co_bb9ad52aa7012efd28900b89f8356405"},{"public_id":"co_be07174c084812cc864458b120e14052","status":"active","name":"pixel-perfect synthetic image training","description":"A training approach that uses precisely generated synthetic images for learning depth estimation.","types":["method"],"aliases":[],"contributors":[{"id":136,"public_id":"3c2apqe3ut","public_label":"Anonymous (3c2apqe3ut)","roles":["extraction"],"url":"https://sah.borca.ai/u/3c2apqe3ut"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/concepts/co_be07174c084812cc864458b120e14052"},{"public_id":"co_c82fc5dc27270652f27e8b2868927a5d","status":"active","name":"depth estimation","description":"The estimation of depth information from visual input in the described task.","types":["task"],"aliases":[],"contributors":[{"id":136,"public_id":"3c2apqe3ut","public_label":"Anonymous (3c2apqe3ut)","roles":["extraction"],"url":"https://sah.borca.ai/u/3c2apqe3ut"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["review"],"url":"https://sah.borca.ai/u/b2adb6bfad"}],"url":"https://sah.borca.ai/concepts/co_c82fc5dc27270652f27e8b2868927a5d"}],"external_ids":{"DOI":"10.1109/ITSC.2019.8916914","ArXiv":null,"PubMed":null,"PubMedCentral":null,"MAG":2990989819,"DBLP":"conf/itsc/GuoASMO19","ACL":null},"open_access":{"is_open_access":false,"pdf_url":null,"landing_url":"https://sah.borca.ai/papers/208629777","source":null,"pdf_url_source":null,"license":null,"reason":"pdf_url_not_indexed"},"reference_availability":{"status":"available","references_indexed":true,"full_text_available":false,"full_text_source":null,"count_basis":"semantic_scholar_metadata","extraction_status":"not_applicable","reason":null},"source":{"provider":"episteme2","base_corpus":"semantic_scholar_dump","freshness_mode":"unknown","basis":["semantic_scholar_metadata","postgres_metadata"],"limits":["paper metadata is based on indexed upstream scholarly datasets","claims and concepts are available only for extracted papers","absence of claims or concepts means no extracted graph data is available in this response"],"status":"available","degraded":false,"degraded_reasons":[],"diagnostics":{"status":"available","degraded":false,"degraded_reasons":[],"metadata_status":"available","graph_status":"available","abstract_status":"available"},"source_flags":1},"paper_id":632040,"paper_uid":"caf9d7ce-e27c-4545-9a68-227728c7946c","canonical_identity":{"paper_id":632040,"paper_uid":"caf9d7ce-e27c-4545-9a68-227728c7946c","identity_status":"available","lookup_basis":"semantic_scholar_external_id","compatibility_path":"corpus_id"},"url":"https://sah.borca.ai/papers/208629777"}