{"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/image-super-resolution-via-feature-augmented","title":"Image Super-resolution via Feature-augmented Random Forest","arxiv_id":"1712.05248","date":"2017-12-14","proceeding":null,"authors":["Hailiang Li","Kin-Man Lam","Miaohui Wang"],"abstract":"Recent random-forest (RF)-based image super-resolution approaches inherit\nsome properties from dictionary-learning-based algorithms, but the\neffectiveness of the properties in RF is overlooked in the literature. In this\npaper, we present a novel feature-augmented random forest (FARF) for image\nsuper-resolution, where the conventional gradient-based features are augmented\nwith gradient magnitudes and different feature recipes are formulated on\ndifferent stages in an RF. The advantages of our method are that, firstly, the\ndictionary-learning-based features are enhanced by adding gradient magnitudes,\nbased on the observation that the non-linear gradient magnitude are with highly\ndiscriminative property. Secondly, generalized locality-sensitive hashing (LSH)\nis used to replace principal component analysis (PCA) for feature\ndimensionality reduction and original high-dimensional features are employed,\ninstead of the compressed ones, for the leaf-nodes' regressors, since\nregressors can benefit from higher dimensional features. This\noriginal-compressed coupled feature sets scheme unifies the unsupervised LSH\nevaluation on both image super-resolution and content-based image retrieval\n(CBIR). Finally, we present a generalized weighted ridge regression (GWRR)\nmodel for the leaf-nodes' regressors. Experiment results on several public\nbenchmark datasets show that our FARF method can achieve an average gain of\nabout 0.3 dB, compared to traditional RF-based methods. Furthermore, a\nfine-tuned FARF model can compare to or (in many cases) outperform some recent\nstateof-the-art deep-learning-based algorithms.","url_abs":"http://arxiv.org/abs/1712.05248v1","url_pdf":"http://arxiv.org/pdf/1712.05248v1.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":"image-super-resolution-via-feature-augmented","repo_url":"https://github.com/HarleyHK/FARF","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[{"task_slug":"content-based-image-retrieval","task_name":"Content-Based Image Retrieval"},{"task_slug":"dictionary-learning","task_name":"Dictionary Learning"},{"task_slug":"dimensionality-reduction","task_name":"Dimensionality Reduction"},{"task_slug":"image-retrieval","task_name":"Image Retrieval"},{"task_slug":"image-super-resolution","task_name":"Image Super-Resolution"},{"task_slug":"retrieval","task_name":"Retrieval"},{"task_slug":"super-resolution","task_name":"Super-Resolution"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/image-super-resolution-on-bsd100-4x-upscaling","task":"Image Super-Resolution","dataset":"BSD100 - 4x upscaling","model":"FAFR*","rank_in_archive_order":54,"of":71,"metrics":{"PSNR":"26.91"},"uses_additional_data":false},{"leaderboard":"/sota/image-super-resolution-on-set14-4x-upscaling","task":"Image Super-Resolution","dataset":"Set14 - 4x upscaling","model":"FAFR*","rank_in_archive_order":94,"of":104,"metrics":{"PSNR":"27.48"},"uses_additional_data":false}],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}