{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,20]],"date-time":"2026-05-20T22:17:16Z","timestamp":1779315436534,"version":"3.51.4"},"reference-count":45,"publisher":"Association for Computing Machinery (ACM)","issue":"6","license":[{"start":{"date-parts":[[2024,11,19]],"date-time":"2024-11-19T00:00:00Z","timestamp":1731974400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/www.acm.org\/publications\/policies\/copyright_policy#Background"}],"content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":["ACM Trans. Graph."],"published-print":{"date-parts":[[2024,12,19]]},"abstract":"<jats:p>Unbiased Monte Carlo path tracing that is extensively used in realistic rendering produces undesirable noise, especially with low samples per pixel (spp). Recently, several methods have coped with this problem by importing unbiased noisy images and auxiliary features to neural networks to either predict a fixed-sized kernel for convolution or directly predict the denoised result. Since it is impossible to produce arbitrarily high spp images as the training dataset, the network-based denoising fails to produce high-quality images under high spp. More specifically, network-based denoising is inconsistent and does not converge to the ground truth as the sampling rate increases. On the other hand, the post-correction estimators yield a blending coefficient for a pair of biased and unbiased images influenced by image errors or variances to ensure the consistency of the denoised image. As the sampling rate increases, the blending coefficient of the unbiased image converges to 1, that is, using the unbiased image as the denoised results. However, these estimators usually produce artifacts due to the difficulty of accurately predicting image errors or variances with low spp. To address the above problems, we take advantage of both kernel-predicting methods and post-correction denoisers. A novel kernel-based denoiser is proposed based on distribution-free kernel regression consistency theory, which does not explicitly combine the biased and unbiased results but constrain the kernel bandwidth to produce consistent results under high spp. Meanwhile, our kernel regression method explores bandwidth optimization in the robust auxiliary feature space instead of the noisy image space. This leads to consistent high-quality denoising at both low and high spp. Experiment results demonstrate that our method outperforms existing denoisers in accuracy and consistency.<\/jats:p>","DOI":"10.1145\/3687949","type":"journal-article","created":{"date-parts":[[2024,11,19]],"date-time":"2024-11-19T15:46:04Z","timestamp":1732031164000},"page":"1-14","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":3,"title":["Neural Kernel Regression for Consistent Monte Carlo Denoising"],"prefix":"10.1145","volume":"43","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-6326-3209","authenticated-orcid":false,"given":"Qi","family":"Wang","sequence":"first","affiliation":[{"name":"State Key Lab of CAD&amp;CG, Zhejiang University, Hangzhou, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0007-7688-3582","authenticated-orcid":false,"given":"Pengju","family":"Qiao","sequence":"additional","affiliation":[{"name":"Institute of Software Chinese Academy of Sciences, Hangzhou, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3296-7999","authenticated-orcid":false,"given":"Yuchi","family":"Huo","sequence":"additional","affiliation":[{"name":"State Key Lab of CAD&amp;CG, Zhejiang University, Hangzhou, China"},{"name":"Zhejiang Lab, Hangzhou, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0004-3745-2761","authenticated-orcid":false,"given":"Shiji","family":"Zhai","sequence":"additional","affiliation":[{"name":"Institute of Computing Technology, Chinese Academy of Sciences, Hangzhou, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0001-9086-1351","authenticated-orcid":false,"given":"Zixuan","family":"Xie","sequence":"additional","affiliation":[{"name":"Institute of Computing Technology, Chinese Academy of Sciences, Hangzhou, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2868-1920","authenticated-orcid":false,"given":"Wei","family":"Hua","sequence":"additional","affiliation":[{"name":"Zhejiang Lab, Hangzhou, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2662-0334","authenticated-orcid":false,"given":"Hujun","family":"Bao","sequence":"additional","affiliation":[{"name":"State Key Lab of CAD&amp;CG, Zhejiang University, Hangzhou, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1194-3253","authenticated-orcid":false,"given":"Tao","family":"Liu","sequence":"additional","affiliation":[{"name":"Shanghai Maritime University, College of Transport &amp; Communications, Shanghai, China"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"320","published-online":{"date-parts":[[2024,11,19]]},"reference":[{"key":"e_1_2_1_1_1","doi-asserted-by":"publisher","DOI":"10.1109\/INFOCT.2019.8710893"},{"key":"e_1_2_1_2_1","doi-asserted-by":"publisher","DOI":"10.1145\/3414685.3417847"},{"key":"e_1_2_1_3_1","doi-asserted-by":"publisher","DOI":"10.1145\/3528233.3530730"},{"key":"e_1_2_1_4_1","volume-title":"Input-Dependent Uncorrelated Weighting for Monte Carlo Denoising. In SIGGRAPH Asia 2023 Conference Papers. 1--10","author":"Back Jonghee","year":"2023","unstructured":"Jonghee Back, Binh-Son Hua, Toshiya Hachisuka, and Bochang Moon. 2023. Input-Dependent Uncorrelated Weighting for Monte Carlo Denoising. In SIGGRAPH Asia 2023 Conference Papers. 1--10."},{"key":"e_1_2_1_5_1","doi-asserted-by":"publisher","DOI":"10.1145\/3072959.3073708"},{"key":"e_1_2_1_6_1","unstructured":"Benedikt Bitterli. 2016. Rendering resources. https:\/\/benedikt-bitterli.me\/resources\/."},{"key":"e_1_2_1_7_1","unstructured":"Benedikt Bitterli. 2018. Tungsten Renderer. https:\/\/github.com\/tunabrain\/tungsten\/"},{"key":"e_1_2_1_8_1","doi-asserted-by":"publisher","DOI":"10.5555\/3071773.3071785"},{"key":"e_1_2_1_9_1","doi-asserted-by":"publisher","DOI":"10.1007\/s00371-023-02951-6"},{"key":"e_1_2_1_10_1","volume-title":"Proceedings of the Conference on High Performance Graphics","author":"Dammertz Holger","unstructured":"Holger Dammertz, Daniel Sewtz, Johannes Hanika, and Hendrik P. A. Lensch. 2010. Edge-avoiding \u00c0-Trous wavelet transform for fast global illumination filtering. In Proceedings of the Conference on High Performance Graphics (Saarbrucken, Germany) (HPG '10). Eurographics Association, Goslar, DEU, 67--75."},{"key":"e_1_2_1_11_1","volume-title":"Computer Graphics Forum","author":"Fan Hangming","unstructured":"Hangming Fan, Rui Wang, Yuchi Huo, and Hujun Bao. 2021. Real-time monte carlo denoising with weight sharing kernel prediction network. In Computer Graphics Forum, Vol. 40. Wiley Online Library, 15--27."},{"key":"e_1_2_1_12_1","doi-asserted-by":"publisher","DOI":"10.1111\/cgf.14454"},{"key":"e_1_2_1_13_1","doi-asserted-by":"publisher","DOI":"10.1145\/3306346.3322954"},{"key":"e_1_2_1_14_1","doi-asserted-by":"publisher","DOI":"10.1214\/aos\/1176346815"},{"key":"e_1_2_1_15_1","doi-asserted-by":"publisher","DOI":"10.1145\/3550454.3555496"},{"key":"e_1_2_1_16_1","doi-asserted-by":"publisher","DOI":"10.1145\/3406181"},{"key":"e_1_2_1_17_1","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.243"},{"key":"e_1_2_1_18_1","doi-asserted-by":"publisher","DOI":"10.1007\/s41095-021-0209-9"},{"key":"e_1_2_1_19_1","doi-asserted-by":"publisher","DOI":"10.1145\/3450626.3459793"},{"key":"e_1_2_1_20_1","doi-asserted-by":"publisher","DOI":"10.1145\/15922.15902"},{"key":"e_1_2_1_21_1","doi-asserted-by":"publisher","DOI":"10.1145\/2766977"},{"key":"e_1_2_1_22_1","doi-asserted-by":"publisher","DOI":"10.1145\/3306346.3323038"},{"key":"e_1_2_1_23_1","volume-title":"Kingma and Jimmy Ba","author":"Diederik","year":"2014","unstructured":"Diederik P. Kingma and Jimmy Ba. 2014. Adam: A Method for Stochastic Optimization. CoRR abs\/1412.6980 (2014). https:\/\/api.semanticscholar.org\/CorpusID:6628106"},{"key":"e_1_2_1_24_1","doi-asserted-by":"publisher","DOI":"10.1111\/insr.12039"},{"key":"e_1_2_1_25_1","doi-asserted-by":"publisher","DOI":"10.1145\/2366145.2366213"},{"key":"e_1_2_1_26_1","volume-title":"A Detail Preserving Neural Network Model for Monte Carlo Denoising. Computational Visual Media","author":"Lin Weiheng","year":"2020","unstructured":"Weiheng Lin, Beibei Wang, Lu Wang, and Nicolas Holzschuch. 2020. A Detail Preserving Neural Network Model for Monte Carlo Denoising. Computational Visual Media (2020), 1--12."},{"key":"e_1_2_1_27_1","unstructured":"Xiaoxu Meng Quan Zheng Amitabh Varshney Gurprit Singh and Matthias Zwicker. 2020. Real-time Monte Carlo Denoising with the Neural Bilateral Grid.. In EGSR (DL). 13--24."},{"key":"e_1_2_1_28_1","doi-asserted-by":"publisher","DOI":"10.1137\/1109020"},{"key":"e_1_2_1_29_1","unstructured":"Nidjo123. 2019. Tungsten Renderer. https:\/\/github.com\/Nidjo123\/kpcn"},{"key":"e_1_2_1_30_1","doi-asserted-by":"publisher","DOI":"10.1145\/1618452.1618486"},{"key":"e_1_2_1_31_1","unstructured":"Adam Paszke Sam Gross Soumith Chintala Gregory Chanan Edward Yang Zachary DeVito Zeming Lin Alban Desmaison Luca Antiga and Adam Lerer. 2017. Automatic differentiation in PyTorch. (2017)."},{"key":"e_1_2_1_32_1","volume-title":"U-net: Convolutional networks for biomedical image segmentation. In Medical image computing and computer-assisted intervention-MICCAI 2015: 18th international conference","author":"Ronneberger Olaf","year":"2015","unstructured":"Olaf Ronneberger, Philipp Fischer, and Thomas Brox. 2015. U-net: Convolutional networks for biomedical image segmentation. In Medical image computing and computer-assisted intervention-MICCAI 2015: 18th international conference, Munich, Germany, October 5--9, 2015, proceedings, part III 18. Springer, 234--241."},{"key":"e_1_2_1_33_1","doi-asserted-by":"publisher","DOI":"10.1145\/2366145.2366214"},{"key":"e_1_2_1_34_1","doi-asserted-by":"publisher","DOI":"10.1145\/2167076.2167083"},{"key":"e_1_2_1_35_1","volume-title":"Density estimation for statistics and data analysis","author":"Silverman Bernard W","unstructured":"Bernard W Silverman. 2018. Density estimation for statistics and data analysis. Routledge."},{"key":"e_1_2_1_36_1","volume-title":"Attention is all you need. Advances in Neural Information Processing Systems","author":"Vaswani A","year":"2017","unstructured":"A Vaswani. 2017. Attention is all you need. Advances in Neural Information Processing Systems (2017)."},{"key":"e_1_2_1_37_1","doi-asserted-by":"publisher","DOI":"10.1145\/3197517.3201388"},{"key":"e_1_2_1_38_1","series-title":"Series A (1961-2002) 26, 4","volume-title":"Smooth Regression Analysis. Sankhy\u0101: The Indian Journal of Statistics","author":"Watson Geoffrey S.","year":"1964","unstructured":"Geoffrey S. Watson. 1964. Smooth Regression Analysis. Sankhy\u0101: The Indian Journal of Statistics, Series A (1961-2002) 26, 4 (1964), 359--372."},{"key":"e_1_2_1_39_1","doi-asserted-by":"publisher","DOI":"10.1145\/3283254.3283261"},{"key":"e_1_2_1_40_1","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2020.2999891"},{"key":"e_1_2_1_41_1","doi-asserted-by":"publisher","DOI":"10.1145\/3355089.3356547"},{"key":"e_1_2_1_42_1","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2018.2886005"},{"key":"e_1_2_1_43_1","doi-asserted-by":"publisher","DOI":"10.1145\/3478513.3480565"},{"key":"e_1_2_1_44_1","doi-asserted-by":"publisher","DOI":"10.1145\/3478513.3480510"},{"key":"e_1_2_1_45_1","doi-asserted-by":"publisher","DOI":"10.5555\/2816723.2816781"}],"container-title":["ACM Transactions on Graphics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3687949","content-type":"unspecified","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3687949","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,6,19]],"date-time":"2025-06-19T01:09:57Z","timestamp":1750295397000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3687949"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,11,19]]},"references-count":45,"journal-issue":{"issue":"6","published-print":{"date-parts":[[2024,12,19]]}},"alternative-id":["10.1145\/3687949"],"URL":"https:\/\/doi.org\/10.1145\/3687949","relation":{},"ISSN":["0730-0301","1557-7368"],"issn-type":[{"value":"0730-0301","type":"print"},{"value":"1557-7368","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,11,19]]},"assertion":[{"value":"2024-11-19","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}