ValueError: setting an array element with a sequence. The requested array has an inhomogeneous shape after 1 dimensions. The detected shape was (5,) + inhomogeneous part.

[HardikJain02]

I crash into this error while executing
python test.py --trained_model=craft_mlt_25k.pth --refiner_model=craft_refiner_CTW1500.pth --test_folder=myfolder --refine

Traceback (most recent call last):
File "C:\code\CRAFT-pytorch\test.py", line 165, in
bboxes, polys, score_text = test_net(net, image, args.text_threshold, args.link_threshold, args.low_text, args.cuda, args.poly, refine_net)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\code\CRAFT-pytorch\test.py", line 106, in test_net
polys = craft_utils.adjustResultCoordinates(polys, ratio_w, ratio_h)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\code\CRAFT-pytorch\craft_utils.py", line 251, in adjustResultCoordinates
polys = np.array(polys)
^^^^^^^^^^^^^^^
ValueError: setting an array element with a sequence. The requested array has an inhomogeneous shape after 1 dimensions. The detected shape was (5,) + inhomogeneous part.

[HardikJain02]

GPT fixed the issue for now:

def adjustResultCoordinates(polys, ratio_w, ratio_h, ratio_net=2): if len(polys) > 0: for k in range(len(polys)): if polys[k] is not None: polys[k] *= np.array([ratio_w * ratio_net, ratio_h * ratio_net]) return polys

Any better solution?

[dungviet112]

I changed numpy version to 1.23.0 and it worked for me

[HordesOfGhost]

def adjustResultCoordinates(polys, ratio_w, ratio_h, ratio_net=2): if len(polys) > 0: polys_filtered = [poly for poly in polys if poly is not None] # Added Line polys = np.array(polys_filtered) for k in range(len(polys)): if polys[k] is not None: polys[k] *= (ratio_w * ratio_net, ratio_h * ratio_net) return polys
Replace the above code in site-packages\craft_text_detector\craft_utils.py to solve the inhomogeneous array problem.