<rss xmlns:atom="http://www.w3.org/2005/Atom" version="2.0">
    <channel>
        <title>CenterNet-Based - 分类 - Zhaoylee&#39;s Blogs</title>
        <link>https://zhaoylee.github.io/Blogs_lovelt/categories/centernet-based/</link>
        <description>CenterNet-Based - 分类 - Zhaoylee&#39;s Blogs</description>
        <generator>Hugo -- gohugo.io</generator><language>zh-CN</language><lastBuildDate>Mon, 16 Mar 2026 09:12:18 &#43;0800</lastBuildDate><atom:link href="https://zhaoylee.github.io/Blogs_lovelt/categories/centernet-based/" rel="self" type="application/rss+xml" /><item>
    <title>OCM3D: Object-Centric Monocular 3D Object Detection</title>
    <link>https://zhaoylee.github.io/Blogs_lovelt/posts/ocm3d--object-centric-monocular-3d-object-detection/</link>
    <pubDate>Mon, 16 Mar 2026 09:12:18 &#43;0800</pubDate>
    <author>zhaoylee</author>
    <guid>https://zhaoylee.github.io/Blogs_lovelt/posts/ocm3d--object-centric-monocular-3d-object-detection/</guid>
    <description><![CDATA[<div class="featured-image">
                <img src="/Blogs_lovelt/cover.jpg" referrerpolicy="no-referrer">
            </div><hr>
<blockquote>
<p><strong>🏛️ 会议/期刊</strong>：arxiv<br>
<strong>📅 发表年份</strong>：2021<br>
<strong>💻 开源代码</strong>：<a href="https://github.com/mrsempress/OBMO_GUPNet/blob/main/tools/offline_OBMO.py" target="_blank" rel="noopener noreffer ">OBMO_GUPNet</a><br>
<strong>📄 论文题目</strong>：<a href="https://arxiv.org/pdf/2104.06041" target="_blank" rel="noopener noreffer ">OCM3D: Object-Centric Monocular 3D Object Detection</a></p>
</blockquote>
<hr>
<h3 id="1-文献背景研究目的与核心问题">1. 文献背景、研究目的与核心问题</h3>
<ul>
<li>
<p><strong>研究背景</strong>：单目 3D 目标检测（Monocular 3D Object Detection）是一个高度病态（ill-posed）的问题。主流方法通常依赖纯图像或将其转化为伪激光雷达（Pseudo-LiDAR）点云。然而，前者难以捕捉像素间的 3D 空间几何关系，后者则受困于单目深度估计带来的巨大点云噪声。</p>]]></description>
</item>
<item>
    <title>LR3D: Improving Distant 3D Object Detection Using 2D Box Supervision</title>
    <link>https://zhaoylee.github.io/Blogs_lovelt/posts/lr3d--improving-distant-3d-object-detection-using-2d-box-supervision/</link>
    <pubDate>Sun, 15 Mar 2026 22:23:00 &#43;0800</pubDate>
    <author>zhaoylee</author>
    <guid>https://zhaoylee.github.io/Blogs_lovelt/posts/lr3d--improving-distant-3d-object-detection-using-2d-box-supervision/</guid>
    <description><![CDATA[<div class="featured-image">
                <img src="/Blogs_lovelt/cover.jpg" referrerpolicy="no-referrer">
            </div><hr>
<blockquote>
<p><strong>🏛️ 会议/期刊</strong>：CVPR<br>
<strong>📅 发表年份</strong>：2024<br>
<strong>💻 开源代码</strong>：<a href="%e5%a1%ab%e5%86%99%e4%bd%a0%e7%9a%84URL" rel="">无</a><br>
<strong>📄 论文题目</strong>：<a href="https://openaccess.thecvf.com/content/CVPR2024/papers/Yang_Improving_Distant_3D_Object_Detection_Using_2D_Box_Supervision_CVPR_2024_paper.pdf" target="_blank" rel="noopener noreffer ">Improving Distant 3D Object Detection Using 2D Box Supervision</a></p>
</blockquote>
<hr>
<p>这篇由 NVIDIA 等机构的研究人员发表在 CVPR 2024 的重磅论文 <strong>《Improving Distant 3D Object Detection Using 2D Box Supervision》(简称 LR3D)</strong>，切入了一个目前高阶自动驾驶极其头疼的落地难题：<strong>远距离感知（Long-Range Detection）</strong>。它展示了如何用最廉价的标注，榨取单目视觉在远距离上的极限潜力。</p>]]></description>
</item>
<item>
    <title>PLOT: Pseudo-Labeling via Object Tracking for Monocular 3D Object Detection</title>
    <link>https://zhaoylee.github.io/Blogs_lovelt/posts/plot---pseudo-labeling-via-object-tracking-for-monocular-3d-object-detection/</link>
    <pubDate>Sun, 15 Mar 2026 20:52:51 &#43;0800</pubDate>
    <author>zhaoylee</author>
    <guid>https://zhaoylee.github.io/Blogs_lovelt/posts/plot---pseudo-labeling-via-object-tracking-for-monocular-3d-object-detection/</guid>
    <description><![CDATA[<div class="featured-image">
                <img src="/Blogs_lovelt/cover.jpg" referrerpolicy="no-referrer">
            </div><hr>
<blockquote>
<p><strong>🏛️ 会议/期刊</strong>：ICLR<br>
<strong>📅 发表年份</strong>：2026<br>
<strong>💻 开源代码</strong>：<a href="%e5%a1%ab%e5%86%99%e4%bd%a0%e7%9a%84URL" rel="">无</a><br>
<strong>📄 论文题目</strong>：<a href="https://openreview.net/pdf?id=3knS4J9isg" target="_blank" rel="noopener noreffer ">PLOT: Pseudo-Labeling via Object Tracking for Monocular 3D Object Detection</a></p>
</blockquote>
<hr>
<h3 id="一-背景研究目的与核心问题">一、 背景、研究目的与核心问题</h3>
<ul>
<li>
<p><strong>研究背景：</strong> 单目 3D 目标检测模型极度“吃数据”。然而，人工标注 3D 边界框极其昂贵且耗时，导致目前带 3D 标签的数据集规模很小，严重限制了模型的泛化能力。</p>]]></description>
</item>
<item>
    <title>SpikeSMOKE: Spiking Neural Networks for Monocular 3D Object Detection with Cross-Scale Gated Coding</title>
    <link>https://zhaoylee.github.io/Blogs_lovelt/posts/spikesmoke--spiking-neural-networks-for-monocular-3d-object-detection-with-cross-scale-gated-coding/</link>
    <pubDate>Sun, 15 Mar 2026 19:59:10 &#43;0800</pubDate>
    <author>zhaoylee</author>
    <guid>https://zhaoylee.github.io/Blogs_lovelt/posts/spikesmoke--spiking-neural-networks-for-monocular-3d-object-detection-with-cross-scale-gated-coding/</guid>
    <description><![CDATA[<div class="featured-image">
                <img src="/Blogs_lovelt/cover.jpg" referrerpolicy="no-referrer">
            </div><hr>
<blockquote>
<p><strong>🏛️ 会议/期刊</strong>：CVPR / ICCV / ECCV<br>
<strong>📅 发表年份</strong>：2026<br>
<strong>💻 开源代码</strong>：<a href="%e5%a1%ab%e5%86%99%e4%bd%a0%e7%9a%84URL" rel="">无</a><br>
<strong>📄 论文题目</strong>：<a href="https://arxiv.org/pdf/2506.07737" target="_blank" rel="noopener noreffer ">SpikeSMOKE: Spiking Neural Networks for Monocular 3D Object Detection with Cross-Scale Gated Coding</a></p>
</blockquote>
<hr>
<h3 id="一-背景研究目的与核心问题">一、 背景、研究目的与核心问题</h3>
<ul>
<li>
<p><strong>研究背景：</strong> 在自动驾驶等领域，3D 目标检测是核心技术。其中，“单目 3D 目标检测”由于仅依赖单张图像，硬件成本极低，备受青睐。然而，传统基于人工神经网络（ANNs）的模型计算量大、能耗极高，给边缘计算设备的电池续航和散热带来了巨大压力。</p>]]></description>
</item>
<item>
    <title>MonoCT: Overcoming Monocular 3D Detection Domain Shift with Consistent Teacher Models</title>
    <link>https://zhaoylee.github.io/Blogs_lovelt/posts/monoct-overcoming-monocular-3d-detection-domain-shift-with-consistent-teacher-models/</link>
    <pubDate>Thu, 12 Mar 2026 10:25:13 &#43;0800</pubDate>
    <author>zhaoylee</author>
    <guid>https://zhaoylee.github.io/Blogs_lovelt/posts/monoct-overcoming-monocular-3d-detection-domain-shift-with-consistent-teacher-models/</guid>
    <description><![CDATA[<div class="featured-image">
                <img src="/Blogs_lovelt/cover.jpg" referrerpolicy="no-referrer">
            </div><hr>
<blockquote>
<p><strong>🏛️ 会议/期刊</strong>：ICRA<br>
<strong>📅 发表年份</strong>：2025<br>
<strong>💻 开源代码</strong>：<a href="%e5%a1%ab%e5%86%99%e4%bd%a0%e7%9a%84URL" rel="">GitHub 链接</a><br>
<strong>📄 论文题目</strong>：<a href="https://arxiv.org/abs/2503.13743" target="_blank" rel="noopener noreffer ">MonoCT: Overcoming Monocular 3D Detection Domain Shift with Consistent Teacher Models</a></p>
</blockquote>
<hr>
<h2 id="0-一句话总结-tldr">0. 一句话总结 (TL;DR)</h2>
<p><em>(这篇论文用什么方法，解决了什么问题，达到了什么效果)</em><br>
MonoCT 提出了一种基于<strong>一致性教师模型（Consistent Teacher）的</strong>半监督自适应框架，通过在目标域（Target Domain）引入伪标签一致性约束，有效解决了单目 3D 检测在不同数据集间迁移时的深度估计偏差问题。</p>]]></description>
</item>
</channel>
</rss>
