{"id":3458,"date":"2025-09-16T02:18:38","date_gmt":"2025-09-16T02:18:38","guid":{"rendered":"https:\/\/172-234-197-23.ip.linodeusercontent.com\/?page_id=3458"},"modified":"2025-09-16T02:18:38","modified_gmt":"2025-09-16T02:18:38","slug":"voice-clone-guard-few-shot-deepfake-detection-with-xls-r-embeddings-and-gaussian-process-calibration","status":"publish","type":"page","link":"https:\/\/neurosphere-2.tail52f848.ts.net\/wordpress\/?page_id=3458","title":{"rendered":"Voice Clone Guard: Few-Shot Deepfake Detection with XLS-R Embeddings and Gaussian Process Calibration"},"content":{"rendered":"\n<p class=\"wp-block-paragraph\"><\/p>\n\n\n\n<div data-wp-interactive=\"core\/file\" class=\"wp-block-file\"><object data-wp-bind--hidden=\"!state.hasPdfPreview\" hidden class=\"wp-block-file__embed\" data=\"https:\/\/neurosphere-2.tail52f848.ts.net\/wordpress\/wp-content\/uploads\/2025\/09\/Voice-Clone-Guard-Few-Shot-Deepfake-Detection-Benjamin-J-Gilbert-College-of-the-Mainland-Robotic-Process-Automation.pdf\" type=\"application\/pdf\" style=\"width:100%;height:600px\" aria-label=\"Embed of Voice Clone Guard Few-Shot Deepfake Detection Benjamin J Gilbert College of the Mainland Robotic Process Automation.\"><\/object><a id=\"wp-block-file--media-1e09f88f-fcb3-4cdb-afcb-3d7b9f39979a\" href=\"https:\/\/neurosphere-2.tail52f848.ts.net\/wordpress\/wp-content\/uploads\/2025\/09\/Voice-Clone-Guard-Few-Shot-Deepfake-Detection-Benjamin-J-Gilbert-College-of-the-Mainland-Robotic-Process-Automation.pdf\">Voice Clone Guard Few-Shot Deepfake Detection Benjamin J Gilbert College of the Mainland Robotic Process Automation<\/a><a href=\"https:\/\/neurosphere-2.tail52f848.ts.net\/wordpress\/wp-content\/uploads\/2025\/09\/Voice-Clone-Guard-Few-Shot-Deepfake-Detection-Benjamin-J-Gilbert-College-of-the-Mainland-Robotic-Process-Automation.pdf\" class=\"wp-block-file__button wp-element-button\" download aria-describedby=\"wp-block-file--media-1e09f88f-fcb3-4cdb-afcb-3d7b9f39979a\">Download<\/a><\/div>\n\n\n\n<p class=\"wp-block-paragraph\"><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">We present Voice Clone Guard, a few-shot voice<br>deepfake detector that combines XLS-R embeddings with Gaussian Process (GP) classification for superior detection accuracy and probability calibration. Our approach leverages selfsupervised speech representations from Wav2Vec2-XLS-R and<br>employs GP inference to provide well-calibrated uncertainty<br>estimates. Evaluation across synthetic and real-world datasets<br>demonstrates significant improvements: 95.6% AUC (vs 78.2%<br>MFCC baseline), 4.2% Equal Error Rate (vs 18.5%), and<br>substantially better calibration with Expected Calibration Error<br>of 0.032 (vs 0.127). The method excels in few-shot scenarios,<br>achieving 85.2% accuracy with only 4 examples per class, making<br>it practical for deployment with limited training data.<br>Index Terms\u2014Voice deepfake detection, few-shot learning,<br>XLS-R embeddings, Gaussian processes, probability calibration,<br>speech forensics<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/www.facebook.com\/benjamin.j.gilbert\">Facebook<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/arxiv.org\/pdf\/2303.01639\" target=\"_blank\" rel=\"noreferrer noopener\"><\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/arxiv.org\/pdf\/2303.01639\" target=\"_blank\" rel=\"noreferrer noopener\">[PDF] WESPER: Zero-shot and Realtime Whisper to Normal Voice &#8230; &#8211; arXiv<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/arxiv.org\/pdf\/2303.01639\" target=\"_blank\" rel=\"noreferrer noopener\">It achieves user-independent voice conversion in real time. ABSTRACT. Recognizing whispered speech and converting it to normal speech creates &#8230;<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/arxiv.org\/pdf\/2303.01639\" target=\"_blank\" rel=\"noreferrer noopener\">arxiv.org<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/arxiv.org\/html\/2507.13334v1\" target=\"_blank\" rel=\"noreferrer noopener\"><\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/arxiv.org\/html\/2507.13334v1\" target=\"_blank\" rel=\"noreferrer noopener\">A Survey of Context Engineering for Large Language Models &#8211; arXiv<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/arxiv.org\/html\/2507.13334v1\" target=\"_blank\" rel=\"noreferrer noopener\">This survey introduces Context Engineering, a formal discipline that transcends simple prompt design to encompass the systematic optimization of information &#8230;<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/arxiv.org\/html\/2507.13334v1\" target=\"_blank\" rel=\"noreferrer noopener\">arxiv.org<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/arxiv.org\/html\/2505.18318v1\" target=\"_blank\" rel=\"noreferrer noopener\"><\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/arxiv.org\/html\/2505.18318v1\" target=\"_blank\" rel=\"noreferrer noopener\">The Relational Origins of Rules in Online Communities &#8211; arXiv<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/arxiv.org\/html\/2505.18318v1\" target=\"_blank\" rel=\"noreferrer noopener\">One of the original questions motivating our study was \u201cWhy do so many online communities have similar rules?\u201d In the previous section, we &#8230;<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/arxiv.org\/html\/2505.18318v1\" target=\"_blank\" rel=\"noreferrer noopener\">arxiv.org<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/arxiv.org\/pdf\/2402.01662\" target=\"_blank\" rel=\"noreferrer noopener\"><\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/arxiv.org\/pdf\/2402.01662\" target=\"_blank\" rel=\"noreferrer noopener\">[PDF] arXiv:2402.01662v4 [cs.CY] 12 Dec 2024<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/arxiv.org\/pdf\/2402.01662\" target=\"_blank\" rel=\"noreferrer noopener\">Another key expansion to the notion of griefbots is that generative ghosts might exist pre-mortem as a generative clone (an AI agent &#8230;<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/arxiv.org\/pdf\/2402.01662\" target=\"_blank\" rel=\"noreferrer noopener\">arxiv.org<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/arxiv.org\/html\/2506.14018\" target=\"_blank\" rel=\"noreferrer noopener\"><\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/arxiv.org\/html\/2506.14018\" target=\"_blank\" rel=\"noreferrer noopener\">Understanding Content Moderation Policies and User Experiences &#8230;<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/arxiv.org\/html\/2506.14018\" target=\"_blank\" rel=\"noreferrer noopener\">AI and security researchers use two automated methods to ensure the safety of GAI model output: internal model fine-tuning and external content &#8230;<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/arxiv.org\/html\/2506.14018\" target=\"_blank\" rel=\"noreferrer noopener\">arxiv.org<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/www.politifact.com\/article\/2024\/mar\/20\/ai-generated-audio-deepfakes-increasing-we-tested\/\" target=\"_blank\" rel=\"noreferrer noopener\"><\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/www.politifact.com\/article\/2024\/mar\/20\/ai-generated-audio-deepfakes-increasing-we-tested\/\" target=\"_blank\" rel=\"noreferrer noopener\">AI-generated audio deepfakes are increasing. We tested 4 tools &#8230;<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/www.politifact.com\/article\/2024\/mar\/20\/ai-generated-audio-deepfakes-increasing-we-tested\/\" target=\"_blank\" rel=\"noreferrer noopener\">We tested four free online tools that claim to determine whether an audio clip is AI-generated. Only one of them signaled that the Biden-like robocall was &#8230;<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/www.politifact.com\/article\/2024\/mar\/20\/ai-generated-audio-deepfakes-increasing-we-tested\/\" target=\"_blank\" rel=\"noreferrer noopener\">politifact.com<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/github.com\/Trustworthy-AI-Group\/Adversarial_Examples_Papers\" target=\"_blank\" rel=\"noreferrer noopener\"><\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/github.com\/Trustworthy-AI-Group\/Adversarial_Examples_Papers\" target=\"_blank\" rel=\"noreferrer noopener\">Trustworthy-AI-Group\/Adversarial_Examples_Papers: A list &#8230; &#8211; GitHub<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/github.com\/Trustworthy-AI-Group\/Adversarial_Examples_Papers\" target=\"_blank\" rel=\"noreferrer noopener\">A complete list of papers about adversarial examples. It appears that the List of All Adversarial Example Papers has been experiencing crashes over the past &#8230;<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/github.com\/Trustworthy-AI-Group\/Adversarial_Examples_Papers\" target=\"_blank\" rel=\"noreferrer noopener\">github.com<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/ptacts.uspto.gov\/ptacts\/public-informations\/petitions\/1553742\/download-documents?artifactId=RNRlD0VDtG5tueExPU3AC_K0xf_NlKh8-fUQSY8N9-YjB1su5tPRKp4\" target=\"_blank\" rel=\"noreferrer noopener\"><\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/ptacts.uspto.gov\/ptacts\/public-informations\/petitions\/1553742\/download-documents?artifactId=RNRlD0VDtG5tueExPU3AC_K0xf_NlKh8-fUQSY8N9-YjB1su5tPRKp4\" target=\"_blank\" rel=\"noreferrer noopener\">[PDF] Decl. ISO Patent Owner&#8217;s Mot. to Amend Case No. IPR2023-00693<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/ptacts.uspto.gov\/ptacts\/public-informations\/petitions\/1553742\/download-documents?artifactId=RNRlD0VDtG5tueExPU3AC_K0xf_NlKh8-fUQSY8N9-YjB1su5tPRKp4\" target=\"_blank\" rel=\"noreferrer noopener\">Detecting deepfake audio through linguistic information. &#8230; Benjamin Mood, Debayan Gupta, Henry &#8230; The State of Voice Cloning Technology. Federal &#8230;<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/ptacts.uspto.gov\/ptacts\/public-informations\/petitions\/1553742\/download-documents?artifactId=RNRlD0VDtG5tueExPU3AC_K0xf_NlKh8-fUQSY8N9-YjB1su5tPRKp4\" target=\"_blank\" rel=\"noreferrer noopener\">ptacts.uspto.gov<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/papers.ssrn.com\/sol3\/Delivery.cfm\/5219689.pdf?abstractid=5219689&amp;mirid=1&amp;type=2\" target=\"_blank\" rel=\"noreferrer noopener\"><\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/papers.ssrn.com\/sol3\/Delivery.cfm\/5219689.pdf?abstractid=5219689&amp;mirid=1&amp;type=2\" target=\"_blank\" rel=\"noreferrer noopener\">[PDF] ABSTRACT &#8230;&#8230;&#8230;&#8230;&#8230;&#8230;&#8230;&#8230;&#8230;&#8230;&#8230;&#8230;&#8230;&#8230;&#8230;&#8230;&#8230;&#8230;&#8230;&#8230;.. &#8211; SSRN<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/papers.ssrn.com\/sol3\/Delivery.cfm\/5219689.pdf?abstractid=5219689&amp;mirid=1&amp;type=2\" target=\"_blank\" rel=\"noreferrer noopener\">that deepfake porn constituted 98 percent of all deepfake videos &#8230; Benjamin S. Sheffner, Senior &#8230; from the sound of the voice or simulation of the voice,.<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/papers.ssrn.com\/sol3\/Delivery.cfm\/5219689.pdf?abstractid=5219689&amp;mirid=1&amp;type=2\" target=\"_blank\" rel=\"noreferrer noopener\">papers.ssrn.com<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/orcid.org\/0009-0006-2298-6538\" target=\"_blank\" rel=\"noreferrer noopener\"><\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/orcid.org\/0009-0006-2298-6538\" target=\"_blank\" rel=\"noreferrer noopener\">ORCID<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/orcid.org\/0009-0006-2298-6538\" target=\"_blank\" rel=\"noreferrer noopener\">ORCID Please enable JavaScript to continue using this application.<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/orcid.org\/0009-0006-2298-6538\" target=\"_blank\" rel=\"noreferrer noopener\">orcid.org<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/github.com\/CorentinJ\/Real-Time-Voice-Cloning\" target=\"_blank\" rel=\"noreferrer noopener\"><\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/github.com\/CorentinJ\/Real-Time-Voice-Cloning\" target=\"_blank\" rel=\"noreferrer noopener\">CorentinJ\/Real-Time-Voice-Cloning &#8211; GitHub<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/github.com\/CorentinJ\/Real-Time-Voice-Cloning\" target=\"_blank\" rel=\"noreferrer noopener\">SV2TTS is a deep learning framework in three stages. In the first stage, one creates a digital representation of a voice from a few seconds of audio. In the &#8230;<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/github.com\/CorentinJ\/Real-Time-Voice-Cloning\" target=\"_blank\" rel=\"noreferrer noopener\">github.com<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/www.reddit.com\/r\/LocalLLaMA\/comments\/1gj14oa\/best_open_source_voice_cloning_if_you_have_lots\/\" target=\"_blank\" rel=\"noreferrer noopener\"><\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/www.reddit.com\/r\/LocalLLaMA\/comments\/1gj14oa\/best_open_source_voice_cloning_if_you_have_lots\/\" target=\"_blank\" rel=\"noreferrer noopener\">Best Open Source Voice Cloning if you have lots of reference audio?<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/www.reddit.com\/r\/LocalLLaMA\/comments\/1gj14oa\/best_open_source_voice_cloning_if_you_have_lots\/\" target=\"_blank\" rel=\"noreferrer noopener\">I&#8217;ve been using ElevenLabs for awhile but now want to self-host. I was really impressed with F5-TTS for its ability to clone using only a few seconds of audio.<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/www.reddit.com\/r\/LocalLLaMA\/comments\/1gj14oa\/best_open_source_voice_cloning_if_you_have_lots\/\" target=\"_blank\" rel=\"noreferrer noopener\">reddit.com<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/github.com\/topics\/xlsr\" target=\"_blank\" rel=\"noreferrer noopener\"><\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/github.com\/topics\/xlsr\" target=\"_blank\" rel=\"noreferrer noopener\">xlsr \u00b7 GitHub Topics<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/github.com\/topics\/xlsr\" target=\"_blank\" rel=\"noreferrer noopener\">The official implementation of the method discussed in the paper Improving Spoken Language Identification with Map-Mix(work accepted at ICASSP-2023).<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/github.com\/topics\/xlsr\" target=\"_blank\" rel=\"noreferrer noopener\">github.com<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/github.com\/cvlab-kaist\/GaussianTalker\" target=\"_blank\" rel=\"noreferrer noopener\"><\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/github.com\/cvlab-kaist\/GaussianTalker\" target=\"_blank\" rel=\"noreferrer noopener\">GitHub &#8211; cvlab-kaist\/GaussianTalker<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/github.com\/cvlab-kaist\/GaussianTalker\" target=\"_blank\" rel=\"noreferrer noopener\">This is our official implementation of the paper &#8220;GaussianTalker: Real-Time High-Fidelity Talking Head Synthesis with Audio-Driven 3D Gaussian Splatting&#8221;<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/github.com\/cvlab-kaist\/GaussianTalker\" target=\"_blank\" rel=\"noreferrer noopener\">github.com<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/github.com\/topics\/voice-clone\" target=\"_blank\" rel=\"noreferrer noopener\"><\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/github.com\/topics\/voice-clone\" target=\"_blank\" rel=\"noreferrer noopener\">voice-clone \u00b7 GitHub Topics<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/github.com\/topics\/voice-clone\" target=\"_blank\" rel=\"noreferrer noopener\">Transform your speech into your favorite celebrity&#8217;s or your customized voice. Our cutting-edge tool converts text or any audio into your desired voice.<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/github.com\/topics\/voice-clone\" target=\"_blank\" rel=\"noreferrer noopener\">github.com<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<p>We present Voice Clone Guard, a few-shot voicedeepfake detector that combines XLS-R embeddings with Gaussian Process (GP) classification for superior detection accuracy and probability calibration. Our approach leverages selfsupervised speech representations from Wav2Vec2-XLS-R andemploys GP inference to provide well-calibrated uncertaintyestimates. Evaluation across synthetic and real-world datasetsdemonstrates significant improvements: 95.6% AUC (vs 78.2%MFCC baseline), 4.2% Equal&hellip;&nbsp;<\/p>\n","protected":false},"author":2,"featured_media":3460,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"neve_meta_sidebar":"","neve_meta_container":"","neve_meta_enable_content_width":"","neve_meta_content_width":0,"neve_meta_title_alignment":"","neve_meta_author_avatar":"","neve_post_elements_order":"","neve_meta_disable_header":"","neve_meta_disable_footer":"","neve_meta_disable_title":"","footnotes":""},"class_list":["post-3458","page","type-page","status-publish","has-post-thumbnail","hentry"],"_links":{"self":[{"href":"https:\/\/neurosphere-2.tail52f848.ts.net\/wordpress\/index.php?rest_route=\/wp\/v2\/pages\/3458","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/neurosphere-2.tail52f848.ts.net\/wordpress\/index.php?rest_route=\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/neurosphere-2.tail52f848.ts.net\/wordpress\/index.php?rest_route=\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/neurosphere-2.tail52f848.ts.net\/wordpress\/index.php?rest_route=\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/neurosphere-2.tail52f848.ts.net\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=3458"}],"version-history":[{"count":0,"href":"https:\/\/neurosphere-2.tail52f848.ts.net\/wordpress\/index.php?rest_route=\/wp\/v2\/pages\/3458\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/neurosphere-2.tail52f848.ts.net\/wordpress\/index.php?rest_route=\/wp\/v2\/media\/3460"}],"wp:attachment":[{"href":"https:\/\/neurosphere-2.tail52f848.ts.net\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=3458"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}