Deepdataspace

What's New
- fix: preview grounding infos with applied filters；
本次更新
- 修复: 预览 Grounding 信息时考虑已应用的筛选条件；

What's New:
- feature: enable visualization of grounding Informations in image previews；

本次更新:
- 功能：在图像预览中支持 Grounding 信息的可视化；

What's New:
- optimize: speed up dataset loading
- feature: support random visit to dataset
- bugfix: fix dataset filter update logic
- upgrade: security upgrades for Python dependencies

本次更新:
- 优化：数据集加载速度
- 功能：支持数据集随机访问
- 修复：修复数据集筛选栏的更新逻辑
- 升级：升级 Python 包的安全更新

What's New
- Feature: Support annotating keypoints, masks, and polygons for collaborative projects
- Optimize: Speed up importing datasets up to 2 million images to over 10x faster.
- Optimize: Delete deprecated label_confirm field from Object model
- Optimize: Optimize annotation editor interaction details
- Bugfix: Improved visualization for keypoints and bounding boxes

本次更新
- 新增：支持在协同标注中标注 keypoints、mask 和 polygon
- 优化：提升大数据集(2百万图片)的导入速度至少 10 倍
- 优化：删除了 Object 对象中不再使用的 label_confirm 字段
- 优化：优化标注编辑器交互细节
- 修复：优化了 keypoints 和 bounding box 的显示问题

What's New：
- Optimize: Enhance the interaction experience of the annotation editor.
- Optimize: Standardize the logic of model interface calls to optimize the intelligent interaction experience.

本次新增：
- 优化：标注编辑器交互体验
- 优化：规范化模型接口调用逻辑，优化智能交互体验

What's New:
- Feature: Added iVP Detection and iVP Mask for intelligent annotation.
- Optimize: Optimized Mask RLE format, adopting the COCO compression format standard.

本次新增：
- 功能：新增 iVP Detection、iVP Mask 智能标注。
- 优化：优化 Mask Rle 格式，采用 COCO 压缩格式标准。