Text to Figure텍스트 → 이미지poster

Neural Fields Scaling Laws Across Tasks — Poster

A conference poster detailing theoretical and empirical scaling laws for Neural Fields. It compares initialization strategies across Image Regression, Super Resolution, NeRF, and Occupancy Fields, demonstrating parameter efficiency for specific activations like Sine and Gaussian.

논문 컨텍스트

Paper title: From Activation to Initialization: Scaling Insights for Optimizing Neural Fields Abstract: A conference poster detailing theoretical and empirical scaling laws for Neural Fields. It compares initialization strategies across Image Regression, Super Resolution, NeRF, and Occupancy Fields, demonstrating parameter efficiency for specific activations like Sine and Gaussian. Paper body (method & results): <!DOCTYPE html> <html lang="en"> <head> <meta content="text/html; charset=utf-8" http-equiv="content-type"/> <title>From Activation to Initialization: Scaling Insights for Optimizing Neural Fields</title>  <meta content="width=device-width, initial-scale=1, shrink-to-fit=no" name="viewport"/> <link href="https://cdn.jsdelivr.net/npm/bootstrap@5.3.0/dist/css/bootstrap.min.css" rel="stylesheet" type="text/css"/> <link href="/static/browse/0.3.4/css/ar5iv.0.7.9.min.css" rel="stylesheet" type="text/css"/> <link href="/static/browse/0.3.4/css/ar5iv-fonts.0.7.9.min.css" rel="stylesheet" type="text/css"/> <link href="/static/browse/0.3.4/css/latexml_styles.css" rel="stylesheet" type="text/css"/> <script src="https://cdn.jsdelivr.net/npm/bootstrap@5.3.0/dist/js/bootstrap.bundle.min.js"></script> <script src="https://cdnjs.cloudflare.com/ajax/libs/html2canvas/1.3.3/html2canvas.min.js"></script> <script src="/static/browse/0.3.4/js/addons_new.js"></script> <script src="/static/browse/0.3.4/js/feedbackOverlay.js"></script> <base href="/html/2403.19205v1/"/></head> <body> <nav class="ltx_page_navbar"> <nav class="ltx_TOC"> <ol class="ltx_toclist"> <li class="ltx_tocentry ltx_tocentry_section"><a class="ltx_ref" href="https://arxiv.org/html/2403.19205v1#S1" title="In From Activation to Initialization: Scaling Insights for Optimizing Neural Fields"><span class="ltx_text ltx_ref_title"><span class="ltx_tag ltx_tag_ref">1 </span>Introduction</span></a></li> <li class="ltx_tocentry ltx_tocentry_section"><a class="ltx_ref" href="https://arxiv.org/html/2403.19205v1#S2" title="In From Activation to Initialization: Scaling Insights for Optimizing Neural Fields"><span class="ltx_text ltx_ref_title"><span class="ltx_tag ltx_tag_ref">2 </span>Related Work</span></a></li> <li class="ltx_tocentry ltx_tocentry_section"><a class="ltx_ref" href="https://arxiv.org/html/2403.19205v1#S3" title="In From Activation to Initialization: Scaling Insights for Optimizing Neural Fields"><span class="ltx_text ltx_ref_title"><span class="ltx_tag ltx_tag_ref">3 </span>Notation</span></a></li> <li class="ltx_tocentry ltx_tocentry_section"> <a class="ltx_ref" href="https://arxiv.org/html/2403.19205v1#S4" title="In From Activation to Initialization: Scaling Insights for Optimizing Neural Fields"><span class="ltx_text ltx_ref_title"><span class="ltx_tag ltx_tag_ref">4 </span>Theoretical Scaling Laws</span></a> <ol class="ltx_toclist ltx_toclist_section"> <li class="ltx_tocentry ltx_tocentry_subsection"><a class="ltx_ref" href="https://arxiv.org/html/2403.19205v1#S4.SS1" title="In 4 Theoretical Scaling Laws ‣ From Activation to Initialization: Scaling Insights for Optimizing Neural Fields"><span class="ltx_text ltx_ref_title"><span class="ltx_tag ltx_tag_ref">4.1 </span>A scaling law for shallow networks</span></a></li> <li class="ltx_tocentry ltx_tocentry_subsection"><a class="ltx_ref" href="https://arxiv.org/html/2403.19205v1#S4.SS2" title="In 4 Theoretical Scaling Laws ‣ From Activation to Initialization: Scaling Insights for Optimizing Neural Fields"><span class="ltx_text ltx_ref_title"><span class="ltx_tag ltx_tag_ref">4.2 </span>A scaling law for deep networks</span></a></li> <li class="ltx_tocentry ltx_tocentry_subsection"><a class="ltx_ref" href="https://arxiv.org/html/2403.19205v1#S4.SS3" title="In 4 Theoretical Scaling Laws ‣ From Activation to Initialization: Scaling Insights for Optimizing Neural Fields"><span class="ltx_text ltx_ref_title"><span class="ltx_tag ltx_tag_ref">4.3 </span>Analyzing the proof methodology</span></a></li> <li class="ltx_tocentry ltx_tocentry_subsection"><a class="ltx_ref" href="https://arxiv.org/html/2403.19205v1#S4.SS4" title="In 4 Theoretical Scaling Laws ‣ From Activation to Initialization: Scaling Insights for Optimizing Neural Fields"><span class="ltx_text ltx_ref_title"><span class="ltx_tag ltx_tag_ref">4.4 </span>Designing new initializations</span></a></li> </ol> </li> <li class="ltx_tocentry ltx_tocentry_section"> <a class="ltx_ref" href="https://arxiv.org/html/2403.19205v1#S5" title="In From Activation to Initialization: Scaling Insights for Optimizing Neural Fields"><span class="ltx_text ltx_ref_title"><span class="ltx_tag ltx_tag_ref">5 </span>Experiments: Applications to Neural Fields</span></a> <ol class="ltx_toclist ltx_toclist_section"> <li class="ltx_tocentry ltx_tocentry_subsection"> <a class="ltx_ref" href="https://arxiv.org/html/2403.19205v1#S5.SS1" title="In 5 Experiments: Applications to Neural Fields ‣ From Activation to Initialization: Scaling Insights for Optimizing Neural Fields"><span class="ltx_text ltx_ref_title"><span class="ltx_tag ltx_tag_ref">5.1 </span>Practical Validation of the Theoretical Analysis</span></a> <ol class="ltx_toclist ltx_toclist_subsection"> <li class="ltx_tocentry ltx_tocentry_paragraph"><a class="ltx_ref" href="https://arxiv.org/html/2403.19205v1#S5.SS1.SSS0.Px1" title="In 5.1 Practical Validation of the Theoretical Analysis ‣ 5 Experiments: Applications to Neural Fields ‣ From Activation to Initialization: Scaling Insights for Optimizing Neural Fields"><span class="ltx_text ltx_ref_title">Shallow Experiment:</span></a></li> <li class="ltx_tocentry ltx_tocentry_paragraph"><a class="ltx_ref" href="https://arxiv.org/html/2403.19205v1#S5.SS1.SSS0.Px2" title="In 5.1 Practical Validation of the Theoretical Analysis ‣ 5 Experiments: Applications to Neural Fields ‣ From Activation to Initialization: Scaling Insights for Optimizing Neural Fields"><span class="ltx_text ltx_ref_title">Deep Experiment:</span></a></li> </ol> </li> <li class="ltx_tocentry ltx_tocentry_subsection"><a class="ltx_ref" href="https://arxiv.org/html/2403.19205v1#S5.SS2" title="In 5 Experiments: Applications to Neural Fields ‣ From Activation to Initialization: Scaling Insights for Optimizing Neural Fields"><span class="ltx_text ltx_ref_title"><span class="ltx_tag ltx_tag_ref">5.2 </span>Single Image Super Resolution</span></a></li> <li class="ltx_tocentry ltx_tocentry_subsection"><a class="ltx_ref" href="https://arxiv.org/html/2403.19205v1#S5.SS3" title="In 5 Experiments: Applications to Neural Fields ‣ From Activation to Initialization: Scaling Insights for Optimizing Neural Fields"><span class="ltx_text ltx_ref_title"><span class="ltx_tag ltx_tag_ref">5.3 </span>Occupancy Fields</span></a></li> <li class="ltx_tocentry ltx_tocentry_subsection"><a class="ltx_ref" href="https://arxiv.org/html/2403.19205v1#S5.SS4" title="In 5 Experiments: Applications to Neural Fields ‣ From Activation to Initialization: Scaling Insights for Optimizing Neural Fields"><span class="ltx_text ltx_ref_title"><span class="ltx_tag ltx_tag_ref">5.4 </span>Neural Radiance Fields (NeRF)</span></a></li> <li class="ltx_tocentry ltx_tocentry_subsection"><a class="ltx_ref" href="https://arxiv.org/html/2403.19205v1#S5.SS5" title="In 5 Experiments: Applications to Neural Fields ‣ From Activation to Initialization: Scaling Insights for Optimizing Neural Fields"><span class="ltx_text ltx_ref_title"><span class="ltx_tag ltx_tag_ref">5.5 </span>Physics Informed Neural Networks (PINNs)</span></a></li> </ol> </li> <li class="ltx_tocentry ltx_tocentry_section"><a class="ltx_ref" href="https://arxiv.org/html/2403.19205v1#S6" title="In From Activation to Initialization: Scaling Insights for Optimizing Neural Fields"><span class="ltx_text ltx_ref_title"><span class="ltx_tag ltx_tag_ref">6 </span>Conclusion</span></a></li> <li class="ltx_tocentry ltx_tocentry_section"><a class="ltx_ref" href="https://arxiv.org/html/2403.19205v1#S7" title="In From Activation to Initialization: Scaling Insights for Optimizing Neural Fields"><span class="ltx_text ltx_ref_title"><span class="ltx_tag ltx_tag_ref">7 </span>Limitations</span></a></li> </ol></nav> </nav> <div class="ltx_page_main"> <div class="ltx_page_content"> <article class="ltx_document ltx_authors_1line"> <h1 class="ltx_title ltx_title_document">From Activation to Initialization: Scaling Insights for Optimizing Neural Fields</h1> <div class="ltx_authors"> <span class="ltx_creator ltx_role_author"> <span class="ltx_personname">Hemanth Saratchandran<sup class="ltx_sup" id="id2.2.id1"><span class="ltx_text ltx_font_italic" id="id2.2.id1.1">∗</span></sup> <br class="ltx_break"/>Australian Institute of Machine Learning, <br class="ltx_break"/>University of Adelaide, Australia <br class="ltx_break"/> </span></span> <span class="ltx_author_before"> </span><span class="ltx_creator ltx_role_author"> <span class="ltx_personname">Sameera Ramasinghe <br class="ltx_break"/>Amazon, Australia <br class="ltx_break"/> </span></span> <span class="ltx_author_before"> </span><span class="ltx_creator ltx_role_author"> <span class="ltx_personname">Simon Lucey <br class="ltx_break"/>Australian Institute of Machine Learning, <br class="ltx_break"/>University of Adelaide, Australia <br class="ltx_break"/> </span></span> </div> <div class="ltx_abstract"> <h6 class="ltx_title ltx_title_abstract">Abstract</h6> <p class="ltx_p" id="id3.id1">In the realm of computer vision, Neural Fields have gained prominence as a contemporary tool harnessing neural networks for signal representation. Despite the remarkable progress in adapting these networks to solve a variety of problems, the field still lacks a comprehensive theoretical framework. This article aims to address this gap by delving into the intricate interplay between initialization and activation, providing a foundational basis for the robust optimization of Neural Fields. Our theoretical insights reveal a deep-seated connection among network initialization, architectural choices, and the optimization process, emphasizing the need for a holistic approach when designing cutting-edge Neural Fields.</p> </div> <span class="ltx_note ltx_role_footnotetext" id="footnotex1"><sup class="ltx_note_mark">*</sup><span class="ltx_note_outer"><span class="ltx_note_content"><sup class="ltx_note_mark">*</sup><span class="ltx_note_type">footnotetext: </span>hemanth.saratchandran@adelaide.edu.au</span></span></span> <section class="ltx_section" id="S1"> <h2 class="ltx_title ltx_title_section"> <span class="ltx_tag ltx_tag_section">1 </span>Introduction</h2> <div class="ltx_para" id="S1.p1"> <p class="ltx_p" id="S1.p1.1">Neural Fields have emerged as a compelling paradigm leveraging coordinate-based neural networks to achieve a concise and expressive encoding of intricate geometric structures and visual phenomena <cite class="ltx_cite ltx_citemacro_cite">[<a class="ltx_ref" href="https://arxiv.org/html/2403.19205v1#bib.bib19" title=""><span class="ltx_text" style="font-size:90%;">19</span></a>, <a class="ltx_ref" href="https://arxiv.org/html/2403.19205v1#bib.bib6" title=""><span class="ltx_text" style="font-size:90%;">6</span></a>, <a class="ltx_ref" href="https://arxiv.org/html/2403.19205v1#bib.bib5" title=""><span class="ltx_text" style="font-size:90%;">5</span></a>, <a cl

프롬프트 본문

Above I've shared:
(1) the full paper text,
(2) all paper figures labeled by figure number,
(3) the caption for the central poster figure I'm building.

TASK: This is a CONFERENCE POSTER. **NOT** an academic-paper figure.
Style requirements:

  - Multi-section layout with a clear poster structure: large title banner
    at the top with the paper title + author/affiliation strip, then 3-6
    distinct content panels arranged in columns or a grid.
  - Large legible fonts (text must be readable at 2 m viewing distance) —
    headings ≥ 60 pt visual size in the final image.
  - Use colour blocks / panel backgrounds to delineate sections (this is
    what makes it a poster, not a single-figure diagram).
  - Aspect ratio: portrait or landscape rectangle, NOT square.

If your output looks like a standard academic-paper figure (single panel,
no title banner, dense small text, no colour blocks), you've failed the
task. Render the COMPLETE poster, not just the central figure.

Just give me the final poster image.

지금 이 프롬프트 시도하기

생성기에 자동으로 채워진 상태로 열립니다.

이 프롬프트 시도

Neural Fields Scaling Laws Across Tasks — Poster

논문 컨텍스트

프롬프트 본문

지금 이 프롬프트 시도하기

관련 프롬프트

PINN Framework Pipeline for Microscopy Data

DP-FTRL Privacy Amplification Analysis — Conference Poster

NC-TTT: Noise-Contrastive Test-Time Training Poster