init

SCUT-BIP-Lab · Oct 12, 2024 · 8e778e9 · 8e778e9
1 parent 1f2ccaa
commit 8e778e9
Show file tree

Hide file tree

Showing 18 changed files with 688 additions and 0 deletions.
diff --git a/fingernerf.github.io/README.md b/fingernerf.github.io/README.md
@@ -0,0 +1 @@
+# fingernerf.github.io
diff --git a/fingernerf.github.io/assets/images/comparison_with_mvsnerf.png b/fingernerf.github.io/assets/images/comparison_with_mvsnerf.png
diff --git a/fingernerf.github.io/assets/images/dataset_overview_scut_finger_3d.png b/fingernerf.github.io/assets/images/dataset_overview_scut_finger_3d.png
diff --git a/fingernerf.github.io/assets/images/dataset_overview_scut_fingervein_3d.png b/fingernerf.github.io/assets/images/dataset_overview_scut_fingervein_3d.png
diff --git a/fingernerf.github.io/assets/images/dataset_overview_unsw_3d.png b/fingernerf.github.io/assets/images/dataset_overview_unsw_3d.png
diff --git a/fingernerf.github.io/assets/images/finger3d_points.png b/fingernerf.github.io/assets/images/finger3d_points.png
diff --git a/fingernerf.github.io/assets/images/finger3d_points_only.png b/fingernerf.github.io/assets/images/finger3d_points_only.png
diff --git a/fingernerf.github.io/assets/images/fingernerf_formulas.jpg b/fingernerf.github.io/assets/images/fingernerf_formulas.jpg
diff --git a/fingernerf.github.io/assets/images/fingervein3d_points.png b/fingernerf.github.io/assets/images/fingervein3d_points.png
diff --git a/fingernerf.github.io/assets/images/fingervein3d_points_only.png b/fingernerf.github.io/assets/images/fingervein3d_points_only.png
diff --git a/fingernerf.github.io/assets/images/framework_2.png b/fingernerf.github.io/assets/images/framework_2.png
diff --git a/fingernerf.github.io/assets/images/motivation.png b/fingernerf.github.io/assets/images/motivation.png
diff --git a/fingernerf.github.io/assets/images/ray_sampling.png b/fingernerf.github.io/assets/images/ray_sampling.png
diff --git a/fingernerf.github.io/index.html b/fingernerf.github.io/index.html
@@ -0,0 +1,226 @@
+<!DOCTYPE html>
+<head>
+    <meta charset="utf-8" />
+    <title>Improving 3D Finger Traits Recognition via Generalizable Neural Rendering</title>
+	<link rel="icon" type="image/x-icon" href="../assets/css/images/favicon.ico">
+    <meta content="Improving 3D Finger Traits Recognition via Generalizable Neural Rendering" name="description" />
+    <meta content="summary" name="twitter:card" />
+    <meta content="width=device-width, initial-scale=1" name="viewport" />
+    <link href="static/css/template.css" rel="stylesheet" type="text/css" />
+    <link href="static/css/my_style.css" rel="stylesheet" type="text/css">
+    <link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.0/css/all.min.css">
+    <link rel="stylesheet" href="https://cdn.jsdelivr.net/gh/jpswalsh/academicons@1/css/academicons.min.css">
+
+    <script src="https://ajax.googleapis.com/ajax/libs/webfont/1.6.26/webfont.js" type="text/javascript"></script>
+    <script type="text/javascript">
+        WebFont.load({
+            google: {
+                families: ["Lato:100,100italic,300,300italic,400,400italic,700,700italic,900,900italic", "Montserrat:100,100italic,200,200italic,300,300italic,400,400italic,500,500italic,600,600italic,700,700italic,800,800italic,900,900italic", "Ubuntu:300,300italic,400,400italic,500,500italic,700,700italic", "Changa One:400,400italic", "Open Sans:300,300italic,400,400italic,600,600italic,700,700italic,800,800italic", "Varela Round:400", "Bungee Shade:regular", "Roboto:300,regular,500"]
+            }
+        });
+    </script>
+    <script type="text/javascript">
+        ! function (o, c) {
+            var n = c.documentElement,
+                t = " w-mod-";
+            n.className += t + "js", ("ontouchstart" in o || o.DocumentTouch && c instanceof DocumentTouch) && (n.className += t + "touch")
+        }(window, document);
+    </script>
+    <link href="https://fonts.googleapis.com/css?family=Google+Sans|Noto+Sans|Castoro" rel="stylesheet">
+    <script type="text/javascript" src="static/js/zoom.js"></script>
+    <script type="text/javascript" src="static/js/video_comparison.js"></script>
+    <!-- Google tag (gtag.js) -->
+    <script async src="https://www.googletagmanager.com/gtag/js?id=G-MLDP9MKGC8"></script>
+    <script>
+    window.dataLayer = window.dataLayer || [];
+    function gtag(){dataLayer.push(arguments);}
+    gtag('js', new Date());
+
+    gtag('config', 'G-MLDP9MKGC8');
+    </script>
+</head>
+
+<body>
+    <div class="section hero nerf-_v2">
+        <div class="container-2 nerf_header_v2 w-container">
+            <h1 class="nerf_title_v2">Improving 3D Finger Traits Recognition via Generalizable Neural Rendering</h1>
+            <div class="nerf_subheader_v2">IJCV 2024</div>
+            <div class="nerf_subheader_v2">
+                <div>
+                    <a href="" target="_blank" class="nerf_authors_v2">Hongbin Xu<span
+                            class="text-span_nerf"></span></a><sup> 1</sup>,&nbsp;&nbsp;
+                    <a href="" target="_blank" class="nerf_authors_v2">Junduan Huang<span
+                            class="text-span_nerf"></span></a><sup> 1</sup>,&nbsp;&nbsp;
+                    <a href="" target="_blank" class="nerf_authors_v2">Yuer Ma<span
+                            class="text-span_nerf"></span></a><sup> 1</sup>,&nbsp;&nbsp;
+                    <a href="" target="_blank" class="nerf_authors_v2">Zifeng Li<span
+                            class="text-span_nerf"></span></a><sup> 1</sup>
+                    <a href="" target="_blank" class="nerf_authors_v2">Wenxiong Kang<span
+                        class="text-span_nerf"></span></a><sup> 1*</sup>
+                </div>
+                <div>
+                    <h1 class="nerf_affiliation_v2"><sup>1* </sup>School of Automation Science and Engineering, South China University of Technology,<br>
+                        Wushan Road, Guangzhou, 510641, Guangdong.</h1>
+                </div>
+
+                <div class="external-link">
+                    <a class="btn" href="" role="button" target="_blank">
+                        <i class="ai ai-arxiv"></i> Arxiv </a>
+                    <a class="btn" href="paper/Paper_high_res.pdf" role="button" target="_blank">
+                        <i class="fa fa-file-pdf"></i> Paper </a>
+                    <a class="btn" href="" role="button" target="_blank" disabled>
+                        <i class="fa-brands fa-github"></i> Code </a>
+                    <a class="btn btn-large btn-light" href="" role="button" target="_blank" disabled>
+                        <i class="fa-brands fa-youtube"></i> Video </a>
+                </div>
+
+            </div>
+        </div>
+
+    </div>
+
+
+    <div data-anchor="slide1" class="section nerf_section">
+        <div class="w-container grey_container">
+            <h2 class="grey-heading_nerf">Abstract</h2>
+            <p class="paragraph-3 nerf_text nerf_results_text">
+                3D biometric techniques on finger traits have become a new trend and have demonstrated a powerful ability for recognition and anti-counterfeiting. 
+                Existing methods follow an explicit 3D pipeline that reconstructs the models first and then extracts features from 3D models.
+                However, these explicit 3D methods suffer from the following problems: 1) Inevitable information dropping during 3D reconstruction; 2) Tight coupling between specific hardware and algorithm for 3D reconstruction.
+                It leads us to a question: Is it indispensable to reconstruct 3D information explicitly in recognition  tasks?
+                Hence, we consider this problem in an implicit manner, leaving the nerve-wracking 3D reconstruction problem for learnable neural networks with the help of neural radiance fields (NeRFs).
+                We propose FingerNeRF, a novel generalizable NeRF for 3D finger biometrics.
+                To handle the shape-radiance ambiguity problem that may result in incorrect 3D geometry, we aim to involve extra geometric priors based on the correspondence of binary finger traits like fingerprints or finger veins.
+                First, we propose a novel Trait Guided Transformer (TGT) module to enhance the feature correspondence with the guidance of finger traits.
+                Second, we involve extra geometric constraints on the volume rendering loss with the proposed Depth Distillation Loss and Trait Guided Rendering Loss.
+                To evaluate the performance of the proposed method on different modalities, we collect two new datasets: SCUT-Finger-3D with finger images and SCUT-FingerVein-3D with finger vein images.
+                Moreover, we also utilize the UNSW-3D dataset with fingerprint images for evaluation.
+                In experiments, our FingerNeRF can achieve 4.37% EER on SCUT-Finger-3D dataset, 8.12% EER on SCUT-FingerVein-3D dataset, and 2.90% EER on UNSW-3D dataset, showing the superiority of the proposed implicit method in 3D finger biometrics.
+                <br>
+                <!-- <img  src="assets/images/overview.png"> -->
+            </p>
+        </div>
+    </div>
+
+
+    <!-- <div class="white_section_nerf  w-container">
+        <h2 class="grey-heading_nerf">NeRF-based Generated Results</h2>
+        <div class="grid-container-4">
+            <div>
+                <p class="myprompt nerf_text">A dragon-cat hybrid <br>&nbsp</p>
+                <video class="video" id="1" loop playsinline autoPlay muted
+                src="assets/videos/nerf-based/exp_30.mp4" onplay="resizeAndPlay(this)"></video>
+                <canvas class="videoMerge" id="1_merge"></canvas>
+            </div>
+            <div>
+                <p class="myprompt nerf_text">Albert Einstein with grey suit is riding a bicycle</p>
+                <video class="video" id="3" loop playsinline autoPlay muted
+                src="assets/videos/nerf-based/Einstein.mp4" onplay="resizeAndPlay(this)"></video>
+                <canvas class="videoMerge" id="3_merge"></canvas>
+            </div>
+            <div>
+                <p class="myprompt nerf_text">Mini Paris, highly detailed, <br>8K, HD</p>
+                <video class="video" id="2" loop playsinline autoPlay muted
+                src="assets/videos/nerf-based/miniParis.mp4" onplay="resizeAndPlay(this)"></video>
+                <canvas class="videoMerge" id="2_merge"></canvas>
+            </div>
+            <div>
+                <p class="myprompt nerf_text">A 3D model of mini China town, highly detailed, 8K, HD, blender 3d</p>
+                <video class="video" id="0" loop playsinline autoPlay muted
+                src="assets/videos/nerf-based/ChinaTown.mp4" onplay="resizeAndPlay(this)"></video>
+                <canvas class="videoMerge" id="0_merge"></canvas>
+            </div>
+        </div>
+        <div class="grid-container-1">
+            <a class="mybtn" href="nerf-based-gallery_0.html" role="button">
+             More Results </a>
+        </div>
+    </div>
+
+    <div class="white_section_nerf  w-container">
+        <h2 class="grey-heading_nerf">DMTet-based Generated Results</h2>
+        <div class="grid-container-4">
+            <div>
+                <p class="myprompt nerf_text">A boy in mohawk hairstyle, head only, 4K, HD, raw</p>
+                <video class="video" id="10" loop playsinline autoPlay muted
+                src="assets/videos/dmtet-based/2023_10_08/aboyinmohawkhairstyle_appearance.mp4" onplay="resizeAndPlay(this)"></video>
+                <canvas class="videoMerge" id="10_merge"></canvas>
+            </div>
+            <div>
+                <p class="myprompt nerf_text">Fire-breathing Phoenix, mythical bird, engulfed in flames, <br>rebirth and renewal, 3D render, 8K, HD</p>
+                <video class="video" id="11" loop playsinline autoPlay muted
+                src="assets/videos/dmtet-based/Phoenix7_appearance3.mp4" onplay="resizeAndPlay(this)"></video>
+                <canvas class="videoMerge" id="11_merge"></canvas>
+            </div>
+            <div>
+                <p class="myprompt nerf_text">A bulldog wearing a black pirate hat, highly detailed</p>
+                <video class="video" id="12" loop playsinline autoPlay muted
+                src="assets/videos/dmtet-based/dog_appearance8.mp4" onplay="resizeAndPlay(this)"></video>
+                <canvas class="videoMerge" id="12_merge"></canvas>
+            </div>
+            <div>
+                <p class="myprompt nerf_text">A 3D model of mini China town, highly detailed, 8K, HD, blender 3d</p>
+                <video class="video" id="13" loop playsinline autoPlay muted
+                src="assets/videos/dmtet-based/2023_10_05/ChinaTown0_appearance.mp4" onplay="resizeAndPlay(this)"></video>
+                <canvas class="videoMerge" id="13_merge"></canvas>
+            </div>
+        </div>
+
+        <div class="grid-container-1">
+            <a class="mybtn" href="dmtet-based-gallery_0.html" role="button">
+            More Results </a>
+        </div>
+    </div> -->
+
+
+    <div class="white_section_nerf  w-container">
+        <h2 class="grey-heading_nerf">Method Overview</h2>
+        <div class="grid-container-1">
+            <img src="assets/images/framework_2.png">
+            <p>
+
+            </p>
+        </div>
+    </div>
+
+    <div class="white_section_nerf  w-container">
+        <h2 class="grey-heading_nerf">Qualitative Comparison</h2>
+        <div class="grid-container-1">
+            <img src="assets/images/comparison_with_mvsnerf.png">
+            <p>
+                we provide qualitative comparison
+                results between the proposed FingerNeRF and
+                the representative method of generalizable NeRF,
+                MVSNeRF. The rendered images and depth
+                maps on a continuous sequence of camera tra-
+                jectories are shown in the figure. Due to the
+                shape-radiance ambiguity, we can find that the
+                rendered depth map of MVSNeRF tends to be
+                incorrect. The depth values on the edge area and
+                the center area of finger have no clear difference.
+                It does not make sense, considering that the shape
+                of finger is similar to an elliptic cylinder. Whereas
+                our method can render both the reasonable depth
+                maps and images as shown in the figure.
+            </p>
+        </div>
+    </div>
+
+<!-- <div class="white_section_nerf grey_container w-container">
+<h2 class="grey-heading_nerf">BibTeX</h2>
+<div class="bibtex">
+    <pre><code>@article{sweetdreamer,
+author    = {Weiyu Li and Rui Chen and Xuelin Chen and Ping Tan},
+title     = {Improving 3D Finger Traits Recognition via Generalizable Neural Rendering},
+journal   = {arxiv:2310.02596},
+year      = {2023},
+}</code></pre>
+</div>
+</div> -->
+
+</body>
+<footer>
+    This project page is inspired by <a href="https://github.com/nerfies/nerfies.github.io">Nerfies.</a>, © Weiyu Li. All rights reserved.
+</footer>
+
+</html>