Open kameel2311 opened 3 years ago
你好,请问你这个问题解决了吗?我目前也遇到了和你同样的问题。
@kameel2311 @JonathonLuiten Hi, did you find the answer? I have the same problem as you! thanks
Hi! I also had the same problem: my HOTA metric was very low. Later I found out that my ground truth and predictions datasets were on different scales (for each bounding box in the ground truth dataset the corresponding coordinates of bounding box in predictions were twice as much). After fixing this issue, the script produced an adequate HOTA value.
你好!我也遇到了同样的问题:我的 HOTA 指标非常低。后来我发现我的地面实况数据集和预测数据集的尺度不同(对于地面实况数据集中的每个边界框,预测中边界框的相应坐标是其两倍)。修复此问题后,脚本生成了足够的 HOTA 值。
您好,我发现我的边界框的相应坐标也是gt的两倍,这个是怎么解决的呀
@JonathonLuiten
Hey, hope all is fine. Well I am trying to evaluate my online deepsort tracker. The method uses a separate object detector and thus cannot use premade detections for feature extraction and tracking, but rather live input from a detector were my input video is the MOT16-02 Video. Therefore, some frames are missed and thus removed these frames from the Gt file not to account for missed frames, but the missed detections are not removed from the Gt.
The evaluation output values are so low such that the HOTA is 1.6, MOTA is negative 26.23 and IDF1/IDR/IDP/IDTP values are 0. Is this because I am not using the same pre-generated detections ? And is there another evaluation method that doesnt need any precomputed Detections ?