IndexError: index 2 is out of bounds for axis 0 with size 2

[Dashvvood]

Error Description

--> 215     keypoints = np.array([keypoints[i] for i in self.track_ids_last_frame])
    216     scores = np.array([scores[i] for i in self.track_ids_last_frame])
    218 self.bboxes_last_frame = bboxes_current_frame

IndexError: index 2 is out of bounds for axis 0 with size 2

Root Cause Analysis
In the 1st frame, 1 bbox is detected:

• track ids = [0]
• keypoints.shape = (1, X, Y)
• reordering succeeds: keypoints = np.array([keypoints[i] for i in *self*.track_ids_last_frame])

In the 2nd frame, 2 bboxes are detected. During PoseTracker.track_by_iou(), no matches for second box are found, but since the bbox is large enough, a new track ID (1) is assigned:

• track ids = [0, 1]
• keypoints.shape = (2, X, Y)
• reordering succeeds: keypoints = np.array([keypoints[i] for i in *self*.track_ids_last_frame])

In the 3rd frame, 2 bboxes are detected. During PoseTracker.track_by_iou(), no matches for second box are found, but since the bbox is large enough, a new track ID (2) is assigned:

• track ids = [0, 2]
• keypoints.shape = (2, X, Y)
• reordering fails: keypoints = np.array([keypoints[i] for i in *self*.track_ids_last_frame])
• IndexError: index 2 is out of bounds for axis 0 with size 2

Problematic Code Section

# Reorder keypoints and scores according to track_id
keypoints = np.array([keypoints[i] for i in self.track_ids_last_frame])
scores = np.array([scores[i] for i in self.track_ids_last_frame])

Track ID assignment logic appears to be generating IDs that exceed array dimensions

[FaizAhmeddd]

in call
keypoints = np.array([keypoints[i] for i in self.track_ids_last_frame])
~~~~~~~~~^^^
IndexError: index 2 is out of bounds for axis 0 with size 2

I am facing the same issue of index out of bounds error