From 2d50eb40da8279503354e1caa3e4d8ca3095ab22 Mon Sep 17 00:00:00 2001 From: Hunor Laczko Date: Thu, 6 Jul 2023 16:35:36 +0200 Subject: [PATCH] new scripts --- custom_data/clothing_verts_idx.npy | Bin 0 -> 29144 bytes scripts/calculate_error.py | 176 +++++++++++++++++++++++++++++ scripts/get_error.py | 82 ++++++++++++++ utils/checkpoints.py | 8 +- 4 files changed, 264 insertions(+), 2 deletions(-) create mode 100644 custom_data/clothing_verts_idx.npy create mode 100644 scripts/calculate_error.py create mode 100644 scripts/get_error.py diff --git a/custom_data/clothing_verts_idx.npy b/custom_data/clothing_verts_idx.npy new file mode 100644 index 0000000000000000000000000000000000000000..38994e6dae8d534dd5c6ff91e8a522ec707d780b GIT binary patch literal 29144 zcmbW&QxhOP1EA6Q#samNR~`+%2*vOC3L1({>#@cWltMUb~KsoBY4^q77PiYVv>AJ2h|6 zzRCZcC(o2MW46>OQpfKT|NnP{oEqTYn8xXx!LvNa^Sr=|yu{1A!mGT->%766yv4h` z$NLQMzkd(^V{FD@a;9KPrebQQVP zS%M{5ilteGWm%5pS%DQh8VP1%gi*@7+E zimlm(ZP||P*?}F|iJjSnUD=J@*@HdVi@n*0ec6xwIe-H>h=VzVLphAYIf5fOilaG( zV>yoFIe`;7iIX{nQ#p;(IfFAfi?cb0b2*Rmxqu6~h>N*|OSz28xq>UXimSPXYq^f= zxq%zGiJQ5FTe*$fxq~~oi@Ujpd%2JMd4LCbh=+NEM|q6Ld4eZ-il=#oXL*k2d4U&s ziI;hWS9y)sd4o53i??})cX^NZ`G61kh>!V%|M4lG@i|}cC13G1-|#Kp@jXBABR}yo zzwj%+@jHL;Cx7ub|L`vZ1kxu4WFQ7+5C&y124@I{WGIGa7=~pyhGzsuWF$sr6h>t< zMrRDhWGu#J9L8ll#%BU1WFjVJ5+-FbCT9w!WGbd+8m47Bre_9bWF}^27G`BOW@irO zWG?1r9_D2}=4SyGWFZ!25f)`J7H0{TWGR+r8J1-^mS+W4WF=N+6;@?6R%Z>?WG&Wa z9oA(%)@K7YWFt0a6E?yQj^_kUZs!i}!9`5Bn?&kp>0vz>oaI&-}u#{KoJ6!Jqua-~7YB3=mlV8IXY(m_Znn!5Ew& z7?PnFnqe50;TWC~7?F_}nNb*((HNaE7?ZIWn{gPI@fe>8n2?E>m`RwF$(Woen3AcO znrWDp>6o4wn30*7nOT^X*_fRR?oIFqwDn{zmq^EjUixR8sum`k{n%eb5?xRR^5nrpb0>$sj9xRINH=XjnMc#)TQnOAs~*La;b zc$2qyn|FAZ_jsQV_>hnIm{0g0pYj=>^95h>6<_lW-|`*b^8-Kf6F>6{zw#Tu^9O(O z7k~2)|1v-j{bxW1VqgYgPzGaghG0mBVrYh8ScYSGMqornVq`{PR7PWT#$ZgwVr<4? zT*hO3CSXD)Vqzv?QYK?^reI2@Vrr&gTBc)qW?)8UVrFJxR%T;%=3q|dVs7SPUgl$d z7GOaZVqq3xQ5IuymS9PiVriCPS(amYR$xU|Vr5ogRaRql)?iK6Vr|x8UDjiLHef?G zVq-R8Q#NCBwqQ%PVr#ZxTef3+c3?+#VrOdpRbJzD-r!B%;%(mHUEbq;KHx(>;$uGHe|*Yke9jkq$ya>M zH+;)?e9sU3$WQ#tFZ{}H{LUZz$zS}6w8U znTeU1g;|-6*_nemnTxrZhk2Qg`B{JkS%`&Mghg45#aV(SS&F4uhGkifOmghGRL7<2iv7If;`wg;P0=(>a4P zIg7J7hjTfP^SOWvxrmFogiE=M%ejIpxr(c~hHJTw>$!m&xrv*(g=Xrq_d5M>Kg;#lv*Lj0Cd5gDshj)38 z_xXSi`G`;XjL-RzpZJ+y_?6%Ioj>@KzxbPf_?H2K=`#Z|5Cby^gEAO{GXz626hku% z!!jJhGXf(r5+gGTqcR$!GX`Ta7GpCG<1!xOGXWDa5fd{BlQJ2TGX+yJ6;m?}(=r{? zGXpa+6EialvoagAGY4}r7jrWY^D-avvj7XS5DT*ii?SGtvjj`B6ic%V%d#BHvjQu! z5-YO`tFjuavj%Ij7HhK(>#`o}vjH2j5gW4!o3a_3vjtnS69ENM{*QLa}39F9LIA4Cvp-ea|)+&8mDsxXL1&2 za}MWn9_Mob7jh97a|xGn8JBYfS8^3sa}C#W9oKUMH*ymdpRbJzD-r!B% z;%(mHUEbq;KHx(>;$uGN3%=(^e&T0-;a7g+cmCi{{^I{a|Lbl*24Y|aVNeERaE4$= zhGJ-jVOWM^ct&7EMq*?}VN^zAbjDyz#$s&7VO+*zd?sK*CSqbHVNxbza;9KPrebQQ zVOpkRdS+loW@2V$VOC~icIIGC=3;K@)mFN4xjNkU+^Vg@ipJ@E#L7yKky?z@iV{hE5GqOfAA-N@i+hQF9U@CcV7Qv zKn7x924PSJV{nFGNQPo)hGAHSV|YejL`Gs{MqyM&V|2z~OvYkt#$jB>V|*rHLMCEj zCSg)0V{)coN~U6JreRv9V|r#_MrLAWW?@!lV|M0XPUd26=3!puV}2H3K^9_R7GY5q zV{w*XNtR-1mSI_zV|i9!MOI>ER$*0EV|CVGP1a&<)?r=NV|_MYLpEY#HepjXV{^7( zOSWQbwqaYgV|#XBM|NUoc41d`V|VsoPxfMO_F-T4V}B0dKn~(y4&hJ^<8Y4PNRHxY zj^S92<9JTsL{8#lPT^Ee<8;p8OwQtL&f#3n<9sgQLN4NBF5yxx<8rRxO0ME+uHjm) z<9cr3MsDI}ZsAsL<96=gPVVAv?%`hU<9;6CK_22^9^p|Q<8hwgNuJ_qp5a-Z<9S}- zMPA}%Ug1?<<8|KPP2S>d-r-%|<9$BhLq6hTKH-0S%4dAe7ktTAe9bp}%XfUw5B$ha z{LC->%5VJ6ANJnVE%InT^?* zgE^UtxtWJ~nUDEdfCX8Ig;|6}S&YS5f+bmsrCEk$S&rpdffZSam05*VS&h|MgEd); zwONOCS&#MEfDPG*joE}v*^JHEf-TvKt=Wcc*^cemfgRb2o!Nz5*^S-VgFV@cz1fF- z*^m7>fCD**gE@plIgGI)PDwKAO>a-24ye? zX9$L5D28SjhGjU0X9PxMBt~WwMrAZcXAH(JXAb6MF6L$)=4C$SX8{&uAr@v47G*IOX9<>MDVAm# zmSs7XX9ZSdC01q?R%JC-XARb5E!Jio)@41`X9G55BQ|CeHf1w5XA8DuE4F4Ewq-lE zX9spP_$#^-#&mwd(7e8abV$M^iekNm{X{KBvN#_#;W zpZvw&{KLNt5J~?TkbxMOK^T<57@Q#(lA#!yVHlR-7@iRrk&zggQ5coc7@aW~ld%|^ zaTu5J7@rB4kcpU>Ntl$$n4Bq?lBt-QX_%Jjn4TG!k(rp8S(ugCn4LM8lew6id6<{^ zn4bk$kcC*7MOc)@SezwTlBHOhWmuNwSe_MFk(F4PRalkPSe-RkleJizby%16Sf35p zkd4@wP1uyp*qklclC9X9ZP=FW*q$BOk)7C?UD%b~*quGtlfBrReb|@%*q;M9kb^jw zLpYSfIGiImlA}19V>p)MIGz(Yk&`%?Q#h5=IGr;%le0LRb2yjtIG+o+kc+sOOSqKF zxST7vlB>9yYq*x{xSkuhk(;=gTey|mxScz=le@T^d$^bTxSt1jkcW7fM|hOSc$_DA zlBal@XLy$9c%Bz{k(YRxS9q1zc%3(Rlec)AcX*fgc%KjWkdOG7Pxv37@)@7=1z++N zU-J#$@*Usv13&T;Kl2N}@*BVN2Y>PxfAbIjGC*YgXFvvGUV$^He++PU`w`QYqnuqc4Tk%VPE!Re-7Y4 z4&q=A;ZP3aaE{84j-r{ZE;a%S2eLmnrKH_6O;eUL}XMD~V ze92dQ%{P3@cYMze{K!xI%pd&8U;NEK{L27Q^p^n{h=Cb|K^cs}8G<1hilG^XVHu9$ z8G#WQiIEwFQ5lWV8G|tyi?JDpaT$;CnScqIh>4kmNtukvnSv>qim91~X_=1cnSmLZ ziJ6&&S(%O5nS(i*i@BMHd6|#-S%AB^n|rvI`?#M6d6-A|jL(@k>c9Ui33D+&i?SGt zvjjVFK=gm_8OY%r!I2!r(Hz6E9LMpTz=@p1$(+KeoW|*#!I_-J*_^|G!IfOa)m+21T*vj?z>VC*&D_GR+{W$P!JXX2-Q2^y+{gVqz=J%*!#u*H zJjUZZ!IM12(>%koJje6Az>B=Z%e=y?yvFOi!JE9r+q}cOyvO@|z=wRq$9%&7_>|B1 zoG@KzxbPf_?H1<=nn%j5Cby^gEAO{GXz62 z6hku%!!jJhGXf(r5+gGTqcR$!GX`Ta7GpCG<1!xOGXWDa5fd{BlQJ2TGX+yJ6;m?} z(=r{?GXpa+6EialvoagAGY4}r7jrWY^D-avvj7XS5DT*ii?SGtvjj`B6ic%V%d#BH zvjQu!5-YO`tFjuavj%Ij7VEPC8?q4_vk9BB8Jn{OTe1~fvklv_9ow@5JF*iyvkSYj z8@sayd$JdMvk&{SANz9v2XYVxa|nlW7>9ENM{*QLa}39F9LIA4Cvp-ea|)+&8mDsx zXL1&2a}MWn9_Mob7jh97a|xGn8JBYfS8^3sa}C#W9oKUMH*yma literal 0 HcmV?d00001 diff --git a/scripts/calculate_error.py b/scripts/calculate_error.py new file mode 100644 index 0000000..c290dcc --- /dev/null +++ b/scripts/calculate_error.py @@ -0,0 +1,176 @@ +import os +import sys +sys.path.append(os.path.dirname(os.path.abspath(__file__)) + '/..') + +import importlib +import argparse +parser = argparse.ArgumentParser(description='') +parser.add_argument('--name', type=str) +parser.add_argument('--config', type=str) +parser.add_argument('--debug', action='store_true') +parser.add_argument('--iteration', type=str, default='') +parser.add_argument('-g', '--gpu', type=str) +args = vars(parser.parse_args()) + +if args['debug']: + from configs.iccvw.pyr_64_128_cape_f_2 import config +else: + config_module = importlib.import_module(args['config']) + config = config_module.config + +os.environ["CUDA_DEVICE_ORDER"]="PCI_BUS_ID" +os.environ["CUDA_VISIBLE_DEVICES"] = config['gpu'] +if args['debug']: + os.environ["CUDA_LAUNCH_BLOCKING"] = '1' + +import time +import argparse +import wandb +import tqdm +import random +import numpy as np +import plotly.graph_objects as go +import plotly.express as ex +from PIL import Image + + +import torch +import torch.nn as nn +import torch.optim as optim +import torch.nn.functional as F +from tqdm import tqdm +from data.dataset import Dataset +from data.utils import Frame +from data.cloth3d_io import readPC2Frame, readPC2 +from utils.checkpoints import CheckpointSaver, load_model +from torch.utils.data import DataLoader +from utils.utils import text_to_html +from utils.postprocess import fix_collision +from losses.losses import collision_loss, normal_loss, kl_loss +from custom_utils.custom_utils import UV2Mesh, Accumulator, plot_3d, get_summary_html, maskFacesBatch, get_summary_str, plot_3d_collision, group_inputs_for_levels +from plotly.subplots import make_subplots + + + +def calculate(): + errors = [] + + for frame in tqdm(frames): + + # import plotly.graph_objects as go + # mesh_pred = mesh_pred.cpu().numpy() + # mask = batch['mesh_mask'].cpu().numpy() + # i = 3 + # fig = go.Figure(data=[go.Scatter3d(x=mesh_pred[i,mask[i],0], y=mesh_pred[i,mask[i],2], z=mesh_pred[i,mask[i],1], mode='markers', marker=dict(size=1))]) + # fig.update_layout(scene_aspectmode='cube') + # fig.show() + + pred = np.load(f'{out_folder}/{frame}.npz')['arr_0'] + gt = readPC2Frame(f'{config["test_dir"]}/mesh_posed_no_orient/{frame.id}_{frame.garment}.pc16', frame.frame_nr, True) + # mask = np.load(f'{config["test_dir"]}/mesh_mask/{frame.id}_{frame.garment}.npy') + + fix_data = np.load(f'{config["test_dir"]}/mesh_fixed/{frame.id}_{frame.garment}.npz') + mask_convert = fix_data['mas_convert'] + new_mask = fix_data['mask'] + + pred = pred[new_mask] + gt = gt[mask_convert] + + error = np.linalg.norm(pred - gt, axis=1).mean() + errors.append((str(frame), error)) + + # save errors to a csv file + with open(f'{results_folder}/errors/{name}.csv', 'w') as f: + for frame, error in errors: + frame_str = ','.join(str(frame).split('_')) + f.write(f'{frame_str},{error}\n') + + +def calculate_cape(): + from smpl.smpl_torch import SMPL + errors = [] + + clothing_verts_idx = np.load(os.path.join('custom_data', 'clothing_verts_idx.npy')) + cape_mask = np.zeros((6890), bool) + cape_mask[clothing_verts_idx] = True + + smpl = SMPL(path='smpl', orig_smpl=True, device='cpu') + + for frame in tqdm(frames): + + # import plotly.graph_objects as go + # mesh_pred = mesh_pred.cpu().numpy() + # mask = batch['mesh_mask'].cpu().numpy() + # i = 3 + # fig = go.Figure(data=[go.Scatter3d(x=mesh_pred[i,mask[i],0], y=mesh_pred[i,mask[i],2], z=mesh_pred[i,mask[i],1], mode='markers', marker=dict(size=1))]) + # fig.update_layout(scene_aspectmode='cube') + # fig.show() + + if not os.path.exists(f'{out_folder}/{frame}.npz'): + print('missed one') + continue + + + + pred = np.load(f'{out_folder}/{frame}.npz')['arr_0'] + gt = np.load(f'{config["test_dir"]}/mesh_unposed_no_shape/{frame.id}.{frame.frame_nr}.npz')['mesh_unposed_no_shape'] + pose_shape = np.load(f'{config["test_dir"]}/pose_shape/{frame.id}.{frame.frame_nr}.npy').astype(np.float32) + pose, shape = pose_shape[:72], pose_shape[72:] + gender = int(shape[10] >= 0.5) + shape_offsets = torch.tensordot(torch.from_numpy(shape[:10][None, :]), smpl.shapedirs[gender], dims=[[1], [2]])[0,smpl.to_keep,:].numpy() + + gt_shaped = gt + shape_offsets + + min_body = np.load(os.path.join('/data-local/127_2/cape_processed', 'minimal_body_shape', frame.id.split('_')[0], frame.id.split('_')[0] + '_minimal.npy')) + offsets_pred = pred[cape_mask[smpl.to_keep],:] - min_body[clothing_verts_idx,:] + offsets_orig = gt_shaped[cape_mask[smpl.to_keep],:] - min_body[clothing_verts_idx,:] + + error = np.linalg.norm(offsets_pred - offsets_orig, axis=1).mean() + errors.append((str(frame), error)) + + # import plotly.graph_objects as go + # fig = go.Figure(data=[go.Scatter3d(x=pred[:,0], y=pred[:,2], z=pred[:,1], mode='markers', marker=dict(size=1))]) + # fig.add_trace(go.Scatter3d(x=gt[:,0], y=gt[:,2], z=gt[:,1], mode='markers', marker=dict(size=1))) + # fig.update_layout(scene_aspectmode='cube') + # fig.update_layout(scene=dict(xaxis=dict(range=[-1,1]), yaxis=dict(range=[-1,1]), zaxis=dict(range=[-1,1]))) + # fig.show() + + # save errors to a csv file + with open(f'{results_folder}/errors/{name}.csv', 'w') as f: + for frame, error in errors: + split = str(frame).split('_') + if not config['is_cape']: + frame_str = ','.join(split) + else: + id = '_'.join(split[:-2]) + frame_str = f'{id},{split[-2]},{split[-1]}' + f.write(f'{frame_str},{error}\n') + + +if __name__ == '__main__': + skirt_modifier = '_skirt' if config['is_skirt'] else '' + + name = config['name'] + + results_folder = '/data-local/127_2/results' + if args['iteration'] == '': + folders = [f for f in os.listdir(results_folder) if os.path.isdir(os.path.join(results_folder, f)) and f.startswith(config['name'])] + assert len(folders) == 1 + name = folders[0] + else: + name = f'{name}_{args["iteration"]}' + out_folder = f'{results_folder}/{name}' + + os.makedirs(f'{results_folder}/errors', exist_ok=True) + + if config['is_cape']: + from data.dataset_cape import read_frames + frames = read_frames(config['test_dir'], shuffle=False, is_cape=True) + calculate_cape() + else: + frames_file = config['test_frames'] + with open(os.path.join(config['test_dir'], frames_file), 'r') as f: + frames = f.readlines() + frames = [Frame(frames.split()[0], int(frames.split()[1]), frames.split()[2]) for frames in frames] + + calculate() \ No newline at end of file diff --git a/scripts/get_error.py b/scripts/get_error.py new file mode 100644 index 0000000..c03ae10 --- /dev/null +++ b/scripts/get_error.py @@ -0,0 +1,82 @@ +import os +import sys +sys.path.append(os.path.dirname(os.path.abspath(__file__)) + '/..') + +import importlib +import argparse +parser = argparse.ArgumentParser(description='') +parser.add_argument('--name', type=str) +parser.add_argument('--config', type=str) +parser.add_argument('--debug', action='store_true') +parser.add_argument('--iteration', type=str, default='') +parser.add_argument('-g', '--gpu', type=str) +args = vars(parser.parse_args()) + +if args['debug']: + from configs.iccvw.pyr_64_128_cape_f_2 import config +else: + config_module = importlib.import_module(args['config']) + config = config_module.config + + +import time +import argparse +import tqdm +import random +import numpy as np +import plotly.graph_objects as go +import plotly.express as ex +from data.utils import Frame + + + + +def calculate(): + errors = dict() + errors_mean = dict() + + with open(f'{results_folder}/errors/{name}.csv', 'r') as f: + lines = f.readlines() + + for line in lines: + line = line.split(',') + category = line[2] + if category not in errors: + errors[category] = [] + errors[category].append(float(line[3])) + + for category in errors: + errors_mean[category] = np.mean(errors[category]) + + print(errors_mean) + + mean = np.mean(list(errors_mean.values())) + + print(f'Final error: {mean}') + + + +if __name__ == '__main__': + skirt_modifier = '_skirt' if config['is_skirt'] else '' + + name = config['name'] + if not config['is_cape']: + frames_file = config['test_frames'] + with open(os.path.join(config['test_dir'], frames_file), 'r') as f: + frames = f.readlines() + frames = [Frame(frames.split()[0], int(frames.split()[1]), frames.split()[2]) for frames in frames] + else: + from data.dataset_cape import read_frames + frames = read_frames(config['test_dir'], shuffle=False, is_cape=True) + + results_folder = '/data-local/127_2/results' + if args['iteration'] == '': + folders = [f for f in os.listdir(results_folder) if os.path.isdir(os.path.join(results_folder, f)) and f.startswith(config['name'])] + assert len(folders) == 1 + name = folders[0] + else: + name = f'{name}_{args["iteration"]}' + out_folder = f'{results_folder}/{name}' + + + calculate() \ No newline at end of file diff --git a/utils/checkpoints.py b/utils/checkpoints.py index 8d23549..52d54bd 100644 --- a/utils/checkpoints.py +++ b/utils/checkpoints.py @@ -33,12 +33,16 @@ class CheckpointSaver: def load_model(model: torch.nn.Module, checkpoint, optimizer: torch.nn.Module=None, scaler: torch.nn.Module=None, iteration=''): # list of checkpoint files in the folder checkpoints - checkpoints = [os.path.join(checkpoint, f) for f in os.listdir(checkpoint) if f.endswith(f'{iteration}.pt')] + if iteration == '': + checkpoints = [os.path.join(checkpoint, f) for f in os.listdir(checkpoint) if f.endswith('.pt')] + else: + checkpoints = [os.path.join(checkpoint, f) for f in os.listdir(checkpoint) if f.endswith(f'iteration_{iteration}.pt')] + assert len(checkpoints) == 1 iterations = np.asarray([int(f.split('_')[-1].split('.')[0]) for f in checkpoints]) last_iteration = np.argmax(iterations) checkpoint = torch.load(checkpoints[last_iteration]) iteration = iterations[last_iteration] - model.load_state_dict(checkpoint['model']) + model.load_state_dict(checkpoint['model'], strict=False) if optimizer is not None: optimizer.load_state_dict(checkpoint['optimizer'].state_dict()) if scaler is not None and checkpoint['scaler'] is not None: scaler.load_state_dict(checkpoint['scaler']) print(f'Loaded model from {checkpoints[last_iteration]} at iteration {iteration}')