Skill for AI agent capabilities
Scanned 9/11/2026
Install to Claude Code
npx -y skills add hiyenwong/ai_collection --skill eas-snn-event-detection --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Eas Snn Event Detection?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/hiyenwong-eas-snn-event-detection)More formats (shields.io, HTML) on the badges page.
---
name: eas-snn---event-based-detection-with-adaptive-samp
description: Skill for AI agent capabilities
---
# EAS-SNN - Event-based Detection with Adaptive Sampling SNN
## Overview
端到端自适应采样事件相机检测框架。利用循环卷积 SNN 的神经动力学作为时间事件采样器,实现完全端到端可学习的事件驱动检测。适用于运动模糊和低光照场景。
**来源论文:** arXiv:2403.12574 - EAS-SNN: End-to-End Adaptive Sampling and Representation for Event-based Detection with Recurrent Spiking Neural Networks
## 触发词
事件相机检测、EAS-SNN、自适应采样、SNN 检测、event camera、spiking neural network detection、event-based detection、adaptive sampling
## 核心方法
### 关键发现
脉冲神经元的神经动力学与理想时间事件采样器行为高度一致,可用于自适应采样。
### 架构组件
1. **自适应采样模块**:循环卷积 SNN + 时间记忆
2. **Residual Potential Dropout (RPD)**:调节电位分布
3. **Spike-Aware Training (SAT)**:解决性能退化问题
### 优势
- 端到端可学习
- 高动态范围
- 处理运动模糊
- 低光照适应
## 使用场景
### 适用情况
- 事件相机目标检测
- 高速运动场景
- 低光照环境
- 资源受限边缘设备
### 数据要求
- 事件相机数据流
- 时间戳和极性信息
- 标注的目标框
## 实施步骤
1. **事件数据预处理**
- 事件流组织为时间窗口
- 提取时空表示
2. **SNN 采样模块配置**
- 设置循环卷积结构
- 配置时间记忆长度
3. **训练策略**
- 应用 RPD 调节电位
- 使用 SAT 稳定训练
4. **检测推理**
- 端到端前向传播
- 输出检测框
## 技术细节
### 神经动力学采样
- 脉冲发放阈值控制采样时机
- 膜电位累积对应事件积分
- 时间记忆捕获历史信息
### 训练技巧
| 技术 | 作用 |
|------|------|
| RPD | 防止电位过度累积 |
| SAT | 处理脉冲稀疏性梯度问题 |
## 与其他方法对比
| 方法 | 自适应采样 | 端到端 | 能效 |
|------|-----------|-------|------|
| EAS-SNN | ✅ 神经动力学 | ✅ | ✅ 高 |
| 传统 CNN | ❌ 固定 | ✅ | ⚠️ 中 |
| 手工采样 | ⚠️ 规则 | ❌ | ⚠️ 中 |
## 工具使用
- `exec`: 运行 PyTorch/SpikingJelly 实现
- `read`: 查看事件数据配置
- `web_fetch`: 获取论文代码
## 注意事项
- 事件相机数据格式需要转换
- SNN 训练需要特殊技巧(SAT)
- 推理时内存占用较低
## 扩展阅读
- 相关技能:`spikingjelly-framework`(SNN 框架)
- 相关技能:`decolle-snn-learning`(局部学习)
- 论文链接:https://arxiv.org/abs/2403.12574
## Description
EAS-SNN - Event-based Detection with Adaptive Sampling SNN
## Activation Keywords
- eas-snn-event-detection
- eas-snn-event-detection 技能
- eas-snn-event-detection skill
## Tools Used
- `read` - Read documentation and references
- `web_search` - Search for related information
- `web_fetch` - Fetch paper or documentation
## Instructions for Agents
Follow these steps when applying this skill:
### Step 1: 自适应采样模块
### Step 2: Residual Potential Dropout (RPD)
### Step 3: Spike-Aware Training (SAT)
### Step 4: 事件数据预处理
### Step 5: SNN 采样模块配置
## Examples
### Example 1: Basic Application
**User:** I need to apply EAS-SNN - Event-based Detection with Adaptive Sampling SNN to my analysis.
**Agent:** I'll help you apply eas-snn-event-detection. First, let me understand your specific use case...
**Context:** Apply the methodology
### Example 2: Advanced Scenario
**User:** Complex analysis scenario
**Agent:** Based on the methodology, I'll guide you through the advanced application...
### Example 2: Advanced Application
**User:** What are the key considerations for eas-snn-event-detection?
**Agent:** Let me search for the latest research and best practices...

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!