CUDA-PointPillars icon indicating copy to clipboard operation
CUDA-PointPillars copied to clipboard

What should I do if I want more labels?

Open brillint opened this issue 1 year ago • 1 comments
trafficstars

when I have 8 labels,it returns: onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Transpose, node name: Transpose_309): [ShapeInferenceError] Inferred shape and existing shape differ in dimension 3: (128) vs (32)

brillint avatar Dec 22 '23 02:12 brillint

You need to modify the shape of cls_preds , box_preds , dir_cls_preds according to the your grid size and number of categories, in the simplify_postprocess() func.

ybcc2015 avatar Jan 26 '24 06:01 ybcc2015