Badcase visualization and analysis badcase可视化分析

[OpenMMLab-Assistant-004]

Badcase visualization and analysis

badcase可视化分析

Enhance the functionality of tools/test.py to facilitate visual comparison and analysis of predicted results by the model.

1. Support visualizing model predictions and ground truth (gt) in each image with different color schemes (distinguishing elements with basic colors, e.g., using green for ground truth and red for model outputs).
2. Calculate the loss value for each image for further analysis, and support to print it out or export the final results to a JSON file.
3. Enable result filtering based on user-specified loss thresholds, with default behavior visualizing all images.
4. Provide options for both displaying images and saving them locally.

增强 tools/test.py 功能，方便对模型预测不好的结果进行可视化对比和分析。

1. 支持在每张图中可视化模型预测结果和gt，配色方案上进行一定程度的区分便于观察（不需要仔细设计每个点和肢体颜色，比如gt用绿色，model输出全用红色）
2. 计算每张图片的loss值便于分析，支持直接print或最终结果导出为json文件
3. 支持根据用户预设的loss来进行结果过滤，默认情况会对全部图片进行可视化
4. 支持每张图显示，也支持保存到本地

Task Information

Repository：MMPose

Difficulty：Hard

Points：50

Estimated Time Consumption：21 days

Task Status

Claimant's Github ID：Indigo6

Claim Time：2023年07月25日

PR Link：open-mmlab/mmpose#2584

Points to Note

To claim the task, please reply Claim the mission in the comment section. After claiming the task, please add our assistant bot called MeowMewo on WeChat at: openmmlabwx.

After completing the task, please submit a PR with the title starting with [MMSIG]. After submission, please reply with your PR link in the comments section of this Discussion. We will review your work as soon as possible.

在评论区回复 Claim the mission 领取任务，领取任务后请及时添加小助手喵喵微信: openmmlabwx

完成任务后提交相关链接（如是PR，标题请以[MMSIG]开头）回复到本Disucssion评论区，我们会第一时间审核您的贡献。

[Indigo6]

Claim the mission

[Indigo6]

open-mmlab/mmpose#2584