search

sangyun884 / HR-VITON

Official PyTorch implementation for the paper High-Resolution Virtual Try-On with Misalignment and Occlusion-Handled Conditions (ECCV 2022).
831 stars 171 forks source link

Unable to get `image-parse-agnostic-v3.2` version of the test image. #40

Open khurramkhalil opened 1 year ago

khurramkhalil commented 1 year ago

Hi there, Thankyou for posting Preprocessing.md file. It is really helpful. However, get_parse_agnostic.py seems not working for me. I am feeding openpose_json and image-parse-v3 version of test image to the script but the output file is almost same as input. Can you please tell me what's wrong here?

Thanks (files are attached for reference). pose_keypoints_2d json file: {"version":1.3,"people":[{"person_id":[-1],"pose_keypoints_2d":[377.347,150.683,0.878828,377.364,329.162,0.854779,255.485,329.12,0.750042,229.968,578.648,0.784553,193.315,822.293,0.783354,507.766,326.454,0.74983,533.321,575.794,0.799812,550.191,822.278,0.777631,374.589,779.774,0.595498,283.929,779.74,0.538896,0,0,0,0,0,0,465.266,785.431,0.525901,0,0,0,0,0,0,351.767,122.423,0.962565,405.737,122.434,0.886836,315.058,144.963,0.557863,439.813,142.213,0.784597,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],"face_keypoints_2d":[319.38,127.579,0.751224,319.38,143.588,0.809515,320.076,160.294,0.790853,322.164,176.999,0.794225,325.644,193.008,0.849269,335.389,205.537,0.828812,347.222,217.37,0.792474,363.231,227.114,0.842114,379.936,229.899,0.846106,396.641,227.114,0.859212,411.258,217.37,0.772164,423.091,204.841,0.810962,430.051,190.92,0.796223,432.14,174.215,0.775057,433.532,156.813,0.760198,434.228,140.804,0.765868,434.228,125.491,0.735009,329.82,112.266,0.779718,336.085,108.09,0.863761,346.526,106.698,0.852168,356.27,109.482,0.871376,363.927,112.266,0.890085,389.681,113.658,0.827457,397.337,109.482,0.905375,407.778,106.002,0.904971,418.219,106.698,0.850264,425.179,112.962,0.893872,376.456,125.491,0.869189,376.456,137.324,0.860178,375.76,148.461,0.861719,375.76,160.294,0.912359,363.231,165.166,0.939292,369.495,167.254,0.916936,375.76,170.734,0.934891,384.112,167.254,0.867908,391.073,165.166,0.937826,340.957,125.491,0.906943,347.222,121.315,0.950957,355.574,122.011,0.877362,362.535,126.883,0.89843,354.182,127.579,0.881709,346.526,126.883,0.954901,391.769,126.883,0.96355,398.729,122.011,0.855446,407.778,121.315,0.923747,413.346,126.187,0.970141,407.778,126.883,0.95683,399.425,127.579,0.879406,347.222,179.783,0.882068,358.358,176.999,0.948998,369.495,176.303,0.93487,377.152,176.999,0.893785,385.504,176.303,0.888562,397.337,176.303,0.90328,408.474,177.695,0.942415,398.033,193.008,0.870786,386.896,198.576,0.876316,378.544,199.272,0.901641,368.799,199.272,0.939339,357.662,193.008,0.901118,352.094,181.175,0.9361,369.495,181.175,0.871688,377.848,181.871,0.856293,386.2,180.479,0.853609,404.994,179.783,0.859839,386.2,193.008,0.884429,378.544,193.704,0.921926,368.799,193.008,0.933628,352.094,124.099,0.839171,402.906,124.795,0.872935],"hand_left_keypoints_2d":[543.673,834.132,0.782809,527.56,854.272,0.837533,513.462,888.512,0.858695,511.448,920.737,0.801544,508.427,942.892,0.701775,547.701,908.653,0.732341,535.617,945.913,0.794778,519.504,956.99,0.810137,504.399,960.011,0.750657,556.765,912.681,0.752773,544.68,951.955,1.00048,522.525,962.026,0.898601,505.406,962.026,0.746398,559.786,911.674,0.810647,550.722,945.913,0.731902,528.568,955.983,0.856978,512.455,961.018,0.636564,559.786,907.645,0.697951,550.722,935.843,0.632913,536.624,946.92,0.592884,524.539,948.934,0.407364],"hand_right_keypoints_2d":[193.271,833.399,0.763477,209.34,851.477,0.851761,217.374,881.605,0.874549,217.374,912.738,0.922487,221.391,936.841,0.837963,185.237,903.7,0.735815,200.301,936.841,0.788084,214.361,953.914,0.830897,224.404,967.974,0.823097,177.203,904.704,0.738815,191.263,943.871,0.871404,209.34,958.936,0.916983,225.408,969.983,0.95493,176.198,904.704,0.774935,186.241,937.846,0.737773,206.327,953.914,0.80248,221.391,964.961,0.70411,179.211,899.683,0.637081,186.241,928.807,0.55934,200.301,941.863,0.522707,213.357,946.884,0.423154],"pose_keypoints_3d":[],"face_keypoints_3d":[],"hand_left_keypoints_3d":[],"hand_right_keypoints_3d":[]}]} parse-v3 00000_00

output 00000_00

khurramkhalil commented 1 year ago

I also tried blurred version of the parse-v3 image. But has no difference in output.

lujiazho commented 1 year ago

you may use the wrong image. You should find that the segmentation image's pixel value is not 0 - 20. For example, you need the 00568_00.png under cihp_parsing_maps instead of 00568_00_vis.png.