Segformer -encode images with more 3 channels

Hi everyone,
Thank you for the work of HuggingFace Team !

I try to use the model Segformer with images with more 3 channels.

(7, 512, 512)

from transformers import SegformerForSemanticSegmentation, SegformerFeatureExtractor
feature_extractor = SegformerFeatureExtractor(align=False, reduce_zero_label=False)

inputs = feature_extractor(images=image, return_tensors=“pt”)
ValueError: Unable to infer channel dimension format

Anyone for help me ?
Thank you