PoseClassificationNet Model training on custom dataset

Hi Team,

the Pose Classification (in the Tao toolkit) seems to be intended for humans only and supports a maximum of 34 key points.

is it possible to train this on animals with 38 key points?

we have applied the following custom modification to training; will this work and support the training?

elif layout == “animal-38kpts”:
self.num_node = 38
self.num_person = 1
self.seq_length = 150
self_link = [(i, i) for i in range(self.num_node)]
neighbor_1base = [[28, 36], [28, 34], [28, 0], [36, 35], [35, 29], [29, 32], [32, 33], [34, 37], [37, 30], [30, 31], [31, 33],
[0, 14], [14, 19], [19, 20], [20, 21], [21, 22], [14, 15],
[15, 16], [16, 17], [17, 18], [0, 1], [1, 2], [2, 3], [3, 4], [4, 5], [5, 10], [10, 11],
[11, 12], [12, 13], [5, 6], [6, 7], [7, 8], [8, 9], [4, 23],
[23, 24], [24, 25], [25, 26], [26, 27]]
neighbor_link = [(i, j) for (i, j) in neighbor_1base]
self.edge = self_link + neighbor_link
self.center = 0 #Head
# elif layout==“customer settings”
# pass
else:

Kind Regards,
Riddhi

There is no update from you for a period, assuming this is not an issue anymore. Hence we are closing this topic. If need further support, please open a new one. Thanks

Looks ok. You can try to run training.

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.