Skip to content
View lartpang's full-sized avatar
😶‍🌫️
Learning and Working
😶‍🌫️
Learning and Working

Highlights

  • Pro

Block or report lartpang

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
lartpang/README.md

Hi 👋, I'm lartpang

Python & PyTorch developer · Deep-learning practitioner · Open-source contributor

📝 Recent Writing

  • Agent Skills | 技能生成、优化与生命周期管理 - 2026-06-01: 本文围绕四篇关于"智能体技能"的研究论文展开讨论,探讨了在不频繁修改模型权重的情况下,如何将可复用的过程性知识作为外部技能,成为智能体持续改进的主要载体。四篇论...
  • 循环与注意力:多算几步到底值不值 - 2026-05-30: 本文探讨了六篇关于“循环”机制的论文,重点分析了循环计算在深度学习中的新应用形式。这些研究不是简单回归传统RNN,而是在Transformer等并行架构中探索如...
  • 深度隐式层 | 可微优化 - 2026-05-28: 本文重点讨论了将优化问题解作为网络层的实现方法。首先区分了传统参数优化与作为网络组件的优化层,后者通过求解小规模凸优化问题定义层输出。作者以凸优化问题为例,阐述...
  • 深度隐式层 | 深度平衡模型 (Deep Equilibrium, DEQ) - 2026-05-28: 深度平衡(DEQ)模型是一种新兴的隐式层架构,通过寻找神经网络迭代过程的不动点来实现计算。该模型将传统深度网络视为无限层迭代的平衡状态,使用单层即可表达任意深度...
  • 深度隐式层 | 神经常微分方程 - 2026-05-28: 本文介绍了如何使用微分方程求解器构建连续时间或连续深度的神经网络模型。详细说明了构建ODE-Net的关键步骤:定义动力学函数、使用odeint求解器,以及通过v...
  • 深度隐式层 | 隐式函数与自动微分 - 2026-05-27: 本文探讨了不动点求解器的高效微分方法。通过隐函数定理提出了一种避免传统自动微分内存消耗过大的新方法,即利用不动点方程的特性直接计算梯度。文章为优化算法和隐式层的...
  • 深度隐式层 | 从显式层到隐式层 - 2026-05-27: 本文介绍了深度学习中的隐式层概念,与传统显式层不同,隐式层通过指定输出应满足的条件而非直接计算步骤来定义。文章以自注意力层为例说明显式层的典型实现方式,然后引入...
  • 循环推理模型:把思考搬进隐空间 - 2026-05-26: 循环推理模型(RRM)提出了一种新的AI推理范式,将部分计算从显式token序列转移到隐空间中进行。传统大模型通过增加token数量来延长推理,而RRM通过在隐...
  • Welford算法 | 从单一到批次 - 2026-05-22: 介绍了Welford算法及其批处理变体,用于高效计算大数据的全局均值和方差。针对内存有限场景,该算法只需单次遍历数据,不保存历史样本,且数值稳定。
  • ArXiv 2507 | Franca: Nested Matryoshka Clustering for Scalable Visual Representation Learning - 2026-05-08: Franca模型通过创新的嵌套套娃聚类、循环掩码和绝对空间属性剥离技术实现了有效的视觉基础模型。该模型完全开源,使用公开数据集训练,在多项任务上超越DINOv2...

View the archives @ csdn@p_lart.

🚀 Open Source Projects

Live star badges update automatically. Click a badge to open the repository's stargazer page.

Research, Evaluation & Computer Vision

Project Focus Stars
Awesome-Class-Activation-Map Curated papers, methods, and tools for class activation mapping. GitHub stars
PySODEvalToolkit Evaluation toolkit for salient-object detection and video-object segmentation. GitHub stars
PySODMetrics Efficient implementations of commonly used salient-object-detection metrics. GitHub stars
MethodsCmp Compare FLOPs/MACs, parameter counts, and FPS for PyTorch models. GitHub stars
mssim.pytorch Differentiable SSIM and MS-SSIM implementations for PyTorch. GitHub stars
tta.pytorch Lightweight test-time augmentation utilities for PyTorch. GitHub stars
OpticalFlowBasedVOS Codebase for optical-flow-based video-object segmentation. GitHub stars
CoSaliencyProj Research code and resources for co-saliency detection. GitHub stars

Developer Tools & Learning Resources

Project Focus Stars
BibSentry Local-first browser tool for reviewing, correcting, and exporting BibTeX files. GitHub stars
Hands-on-Docker 一份面向实践的中文 Docker 使用指南。 GitHub stars
PyTorchTricks Practical PyTorch patterns, snippets, and implementation notes. GitHub stars
PyLoss Reusable loss-function implementations for deep-learning projects. GitHub stars
RunIt Simple scheduler for running code across different devices. GitHub stars
RegisterIt Flexible registry utilities for deep-learning codebases. GitHub stars
YuQueTools Export and archive your own Yuque articles. GitHub stars
ManageMyAttachments Organize attachments in an Obsidian vault. GitHub stars

💾 Datasets & Benchmarks

Dataset / Benchmark Platform Focus Signals
lartpang/OVCamo GitHub Open-vocabulary camouflaged-object segmentation. GitHub stars
yooweey/AugmentedIRSTD1kTestset Hugging Face Augmented test set for the IRSTD-1k infrared-small-target dataset. Hugging Face likes Hugging Face downloads
yooweey/LaMoFCBench ModelScope Benchmark for large-model feature coding. Collection
Open source, reproducible research, and practical deep-learning engineering.

Pinned Loading

  1. awesome-segmentation-saliency-dataset awesome-segmentation-saliency-dataset Public

    A collection of some datasets for segmentation / saliency detection. Welcome to PR...:smile:

    636 98

  2. PyTorchTricks PyTorchTricks Public

    Some tricks of pytorch... ⭐

    1.2k 123

  3. SAMs-CDConcepts-Eval SAMs-CDConcepts-Eval Public

    Inspiring the Next Generation of Segment Anything Models: Comprehensively Evaluate SAM and SAM 2 with Diverse Prompts Towards Context-Dependent Concepts under Different Scenes

    Python 575 31

  4. PySODMetrics PySODMetrics Public

    PySODMetrics: A Simple and Efficient Implementation of Grayscale/Binary Segmentation Metrics

    Python 449 38

  5. OVCamo OVCamo Public

    (ECCV 2024) Open-Vocabulary Camouflaged Object Segmentation

    Python 278 17

  6. PyIRSTDMetrics PyIRSTDMetrics Public

    [NeurIPS 2025 (D&B)] Rethinking Evaluation of Infrared Small Target Detection

    Python 360 25