How to test my own images？

[6216]

How to test my own images and see the panoptic segmentation results ?

[AI-liu]

import os import torch import torch.nn as nn import argparse import cv2 import numpy as np

from upsnet.config.config import * from upsnet.config.parse_args import parse_args

from upsnet.models import *

from PIL import Image, ImageDraw

def get_pallete():

pallete_raw = np.zeros((256, 3)).astype('uint8')
pallete = np.zeros((256, 3)).astype('uint8')

pallete_raw[5, :] =  [111,  74,   0]
pallete_raw[6, :] =  [ 81,   0,  81]
pallete_raw[7, :] =  [128,  64, 128]
pallete_raw[8, :] =  [244,  35, 232]
pallete_raw[9, :] =  [250, 170, 160]
pallete_raw[10, :] = [230, 150, 140]
pallete_raw[11, :] = [ 70,  70,  70]
pallete_raw[12, :] = [102, 102, 156]
pallete_raw[13, :] = [190, 153, 153]
pallete_raw[14, :] = [180, 165, 180]
pallete_raw[15, :] = [150, 100, 100]
pallete_raw[16, :] = [150, 120,  90]
pallete_raw[17, :] = [153, 153, 153]
pallete_raw[18, :] = [153, 153, 153]
pallete_raw[19, :] = [250, 170,  30]
pallete_raw[20, :] = [220, 220,   0]
pallete_raw[21, :] = [107, 142,  35]
pallete_raw[22, :] = [152, 251, 152]
pallete_raw[23, :] = [ 70, 130, 180]
pallete_raw[24, :] = [220,  20,  60]
pallete_raw[25, :] = [255,   0,   0]
pallete_raw[26, :] = [  0,   0, 142]
pallete_raw[27, :] = [  0,   0,  70]
pallete_raw[28, :] = [  0,  60, 100]
pallete_raw[29, :] = [  0,   0,  90]
pallete_raw[30, :] = [  0,   0, 110]
pallete_raw[31, :] = [  0,  80, 100]
pallete_raw[32, :] = [  0,   0, 230]
pallete_raw[33, :] = [119,  11,  32]

train2regular = [7, 8, 11, 12, 13, 17, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 31, 32, 33]

for i in range(len(train2regular)):
    pallete[i, :] = pallete_raw[train2regular[i], :]

pallete = pallete.reshape(-1)

# return pallete_raw
return pallete

parser = argparse.ArgumentParser() args, rest = parser.parse_known_args() args.cfg = "/home/ai/UPSNet-master/upsnet/experiments/upsnet_resnet50_cityscapes_16gpu.yaml" args.weight_path = "/home/ai/UPSNet-master/model/upsnet_resnet_50_cityscapes_12000.pth" args.eval_only = "Ture" update_config(args.cfg)

test_model = eval("resnet_50_upsnet")().cuda() test_model.load_state_dict(torch.load(args.weight_path))

#print(test_model)

for p in test_model.parameters(): p.requires_grad = False

test_model.eval()

im = cv2.imread("lindau_000000_000019_leftImg8bit.png") im_resize = cv2.resize(im,(2048,1024),interpolation=cv2.INTER_CUBIC) im_resize = im_resize.transpose(2, 0, 1)

im_tensor = torch.from_numpy(im_resize) im_tensor =torch.unsqueeze(im_tensor,0).type(torch.FloatTensor).cuda() print(im_tensor.shape) # torch.Size([1, 3, 1024, 2048])

test_fake_numpy_data = np.random.rand(1,3) data = {'data': im_tensor , 'im_info' : test_fake_numpy_data} print(data['im_info']) output = test_model(data) #print(output) print(output['fcn_outputs'])

pallete = get_pallete() segmentation_result = np.uint8(np.squeeze(np.copy(output['fcn_outputs']))) segmentation_result = Image.fromarray(segmentation_result) segmentation_result.putpalette(pallete) segmentation_result = segmentation_result.resize((im.shape[1],im.shape[0])) segmentation_result.save("hello_result.png")

[qiuhaining]

segmentation_result

import os import torch import torch.nn as nn import argparse import cv2 import numpy as np

from upsnet.config.config import * from upsnet.config.parse_args import parse_args

from upsnet.models import *

from PIL import Image, ImageDraw

def get_pallete():

pallete_raw = np.zeros((256, 3)).astype('uint8')
pallete = np.zeros((256, 3)).astype('uint8')

pallete_raw[5, :] =  [111,  74,   0]
pallete_raw[6, :] =  [ 81,   0,  81]
pallete_raw[7, :] =  [128,  64, 128]
pallete_raw[8, :] =  [244,  35, 232]
pallete_raw[9, :] =  [250, 170, 160]
pallete_raw[10, :] = [230, 150, 140]
pallete_raw[11, :] = [ 70,  70,  70]
pallete_raw[12, :] = [102, 102, 156]
pallete_raw[13, :] = [190, 153, 153]
pallete_raw[14, :] = [180, 165, 180]
pallete_raw[15, :] = [150, 100, 100]
pallete_raw[16, :] = [150, 120,  90]
pallete_raw[17, :] = [153, 153, 153]
pallete_raw[18, :] = [153, 153, 153]
pallete_raw[19, :] = [250, 170,  30]
pallete_raw[20, :] = [220, 220,   0]
pallete_raw[21, :] = [107, 142,  35]
pallete_raw[22, :] = [152, 251, 152]
pallete_raw[23, :] = [ 70, 130, 180]
pallete_raw[24, :] = [220,  20,  60]
pallete_raw[25, :] = [255,   0,   0]
pallete_raw[26, :] = [  0,   0, 142]
pallete_raw[27, :] = [  0,   0,  70]
pallete_raw[28, :] = [  0,  60, 100]
pallete_raw[29, :] = [  0,   0,  90]
pallete_raw[30, :] = [  0,   0, 110]
pallete_raw[31, :] = [  0,  80, 100]
pallete_raw[32, :] = [  0,   0, 230]
pallete_raw[33, :] = [119,  11,  32]

train2regular = [7, 8, 11, 12, 13, 17, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 31, 32, 33]

for i in range(len(train2regular)):
    pallete[i, :] = pallete_raw[train2regular[i], :]

pallete = pallete.reshape(-1)

# return pallete_raw
return pallete

parser = argparse.ArgumentParser() args, rest = parser.parse_known_args() args.cfg = "/home/ai/UPSNet-master/upsnet/experiments/upsnet_resnet50_cityscapes_16gpu.yaml" args.weight_path = "/home/ai/UPSNet-master/model/upsnet_resnet_50_cityscapes_12000.pth" args.eval_only = "Ture" update_config(args.cfg)

test_model = eval("resnet_50_upsnet")().cuda() test_model.load_state_dict(torch.load(args.weight_path))

#print(test_model)

for p in test_model.parameters(): p.requires_grad = False

test_model.eval()

im = cv2.imread("lindau_000000_000019_leftImg8bit.png") im_resize = cv2.resize(im,(2048,1024),interpolation=cv2.INTER_CUBIC) im_resize = im_resize.transpose(2, 0, 1)

im_tensor = torch.from_numpy(im_resize) im_tensor =torch.unsqueeze(im_tensor,0).type(torch.FloatTensor).cuda() print(im_tensor.shape) # torch.Size([1, 3, 1024, 2048])

test_fake_numpy_data = np.random.rand(1,3) data = {'data': im_tensor , 'im_info' : test_fake_numpy_data} print(data['im_info']) output = test_model(data) #print(output) print(output['fcn_outputs'])

pallete = get_pallete() segmentation_result = np.uint8(np.squeeze(np.copy(output['fcn_outputs']))) segmentation_result = Image.fromarray(segmentation_result) segmentation_result.putpalette(pallete) segmentation_result = segmentation_result.resize((im.shape[1],im.shape[0])) segmentation_result.save("hello_result.png")

it‘s just the segmentation_result, not panoptic result

[mumuyanyan]

it‘s just the segmentation_result, not panoptic result

yes, the result is semantic segmentation, do you have the panoptic inference?

[discretecoder]

I tried using the panoptic_outputs and using the palette function but it skips the instances somehow

[DdeGeus]

I tried using the panoptic_outputs and using the palette function but it skips the instances somehow

It appears that the 'im_info' entry of the input data dictionary should contain the dimensions of the input image, as it is used for scaling the bounding boxes. So, for Cityscapes, take

im_info = np.array([[1024, 2048, 3]])
data = {'data': im_tensor , 'im_info' : im_info}

instead of the random values from np.random.rand(1,3)

This solved the problem for me. It now outputs panoptic segmentation results including things instances.