kekezack

bbbbcf23f8 chore: update project files
b6152fe0cd feat(rss): 引入 guid 去重机制，增强空标题防护与翻译质量，新增 Markdown 导出，升级至 v6.7.0 - RSS 存储新增 guid 字段，去重优先级改为 guid > url - 解析器/渲染层多级空标题兜底（摘要→URL→feed名称） - 翻译提示词要求保留编号顺序，空翻译不覆盖原始标题 - 防止 URL 格式标题覆盖有意义的已有标题 - HTML 报告新增 Markdown 格式导出（#1121） - 修正 README MCP badge 版本号（v4.0.2 → v4.0.4）
b109701b42 fix: 过滤 platforms.sources 中 enabled: false 的数据源，修复全局开关失效，升级至 v6.6.2/mcp-v4.0.4
ddd3f4dcaf fix(core): 调度器时间区间改为半开区间，修复相邻时段误判重叠
6d82f2fcc8 feat(ci): Issue Guard 增强为两层防护，覆盖 Issue/评论/PR 全场景

hai 1 día

hai 1 día

hai 1 día

7a65a1d472 chore(project): 初始化项目配置文件和代码结构 - 添加 Babel 生产环境配置用于剥离 invariant - 配置 Clang 格式化规则 - 设置 Docker、ESLint、Git 和 Prettier 忽略规则 - 添加 Watchman 配置文件 - 创建多个模块的初始化文件包括 SAM2、SwinTransformer 和工具模块 - 添加前端注释组件和对象添加按钮 - 实现掩码数据处理工具类 - 构建后端 Flask 应用服务 - 配置前端样式表和 Tailwind CSS
235cfeb50c Initial commit
Ver comparación destas 2 achegas »

hai 1 mes

hai 1 mes

99604bec0a docs(readme): Update README.md with detailed installation and usage guide - Changed default prediction result file suffix from _v4_minute to _minute - Added project overview and key features description - Added detailed installation steps and prerequisites - Provided dataset preparation directory structure example - Added training and evaluation command-line parameter tables - Improved model architecture and output structure documentation - Enhanced SwanLab monitoring feature introduction - Added advanced usage including checkpoint resume and component disable options

hai 1 mes

d5359a8c2e feat(Initial commit): XNet - Polyp Segmentation with Wavelet-FFT Enhanced SwinUNETR Features: - Implementation of Wavelet_FFT_SwinUNETR architecture for medical image segmentation - Adaptive Wavelet Augmented Enhancer (AWAE) module for multi-scale feature enhancement - FFT-based frequency domain enhancement via FFTRCAB modules - Combined Dice + CrossEntropy + IoU loss function for robust training - Comprehensive data augmentation pipeline with geometric and photometric transforms - Early stopping mechanism with automatic resume from best checkpoint - SwanLab experiment tracking integration - Full evaluation pipeline with metrics: mDice, mIoU, mHD, mHD95 Project Structure: - train.py: Main training script with configurable hyperparameters - eval.py: Model evaluation and visualization tool - lib/model/: Core model architecture (SwinUNETR backbone) - lib/modules/: Enhancement modules (AWAE, FFTRCAB) - datasets/: Polyp detection dataset loader with MONAI integration Technical Highlights: - Multi-level wavelet decomposition with adaptive attention - Frequency-domain feature enhancement using FFT

hai 1 mes

hai 1 mes

hai 1 mes