def id_to_rgb(cls, id):
"""
Cover float[-2**20~2**20] to float32[0.~1.] rgb
Parameters
----------
id : int, float or numpy
float[-2**20~2**20].
Returns
-------
(3,)float32
RGB for Blender.
"""
if isinstance(id, (float, int)):
id = np.array(id)
rgb = np.zeros(id.shape + (3, ), dtype=np.float64)
rgb[..., 0] = id < 0
absf = np.abs(id)
int_part = absf // 1
# rgb[..., 1] = 1 - 1 / (int_part + 1)
poww = np.int32(np.log2(int_part + 1, dtype=np.float32)) + 1
denominator = (2**poww).round()
rgb[..., 1] = ((int_part - denominator // 2 + 1) * 2 + 1) / denominator
rgb[..., 2] = absf % 1
return rgb
def id_to_rgb(f):
if isinstance(f, (float, int)):
f = np.array(f)
rgb = np.zeros(f.shape + (3, ), dtype=np.float64)
rgb[..., 0] = f < 0
absf = np.abs(f)
int_part = absf // 1
rgb[..., 1] = 1 - 1 / (int_part + 1)
rgb[..., 2] = absf % 1
return rgb
def getMaskOfPgt(argkv):
pgt, cycle_tmpl, shape, out_cyc_r = argkv.get('pgt'), argkv.get(
'cycle_tmpl'), argkv.get('shape'), argkv.get('out_cyc_r')
if not isinstance(pgt, Vector):
pgt = Vector(pgt)
bg = np.zeros(shape[-2:], dtype=np.int32)
if out_cyc_r:
bg -= 1
mask = pasteImg(bg, cycle_tmpl, pgt - cycle_tmpl.shape[-1] // 2)
# g()
# print(pgt)
# show-mask
# tree-argkv
return (mask == 0), (mask == 1),
def __init__(self,
cyc_r=7,
log_freq=60 * 5,
poolings=['avg'],
number_worker=8,
out_cyc_r=None,
layerNorma=True,
probMargin=None,
temper=1):
super(SpatialSoftmax, self).__init__()
self.out_cyc_r = out_cyc_r
self.poolings = poolings
self.log_freq = log_freq
self.cyc_r = cyc_r
self.layerNorma = layerNorma
self.probMargin = probMargin
self.temper = temper
# raw_cycle_tmpl = getWeightCore(300, seta=.25)
# raw_cycle_tmpl = raw_cycle_tmpl > raw_cycle_tmpl[150,0]
# cycle_tmpl_np = resize(raw_cycle_tmpl, np.int8((cyc_r*2, cyc_r*2, )))
cyc_tmpl_np = getCycle(cyc_r)
cycle_tmpl = np.int32(cyc_tmpl_np > .5)
# g()
if out_cyc_r is not None:
# out_cycle_tmpl_np = resize(raw_cycle_tmpl, (out_cyc_r*2, out_cyc_r*2, ))
# out_cycle_tmpl = np.int32(out_cycle_tmpl_np >.5) - 1
out_cycle_tmpl_np = getCycle(out_cyc_r)
out_cycle_tmpl = np.int32(out_cycle_tmpl_np > .5) - 1
inCycIndx = np.zeros(out_cycle_tmpl.shape, np.bool)
pasteImg(inCycIndx, cycle_tmpl, Vector([out_cyc_r - cyc_r] * 2))
out_cycle_tmpl[inCycIndx] = 1
cycle_tmpl = out_cycle_tmpl
# self.cycle_tmpl = th.from_numpy(cycle_tmpl)
self.cycle_tmpl = (cycle_tmpl)
@author: yl
"""
import cv2
from boxx import *
import skimage.data as sda
from boxx import sin, cos
from boxx import Vector, np, timegap, intround
import time
img = sda.astronaut()
imgs = [img]
hw = Vector([800, 800])
imgHw = hw
bg = np.zeros(list(imgHw) + [3]).astype(np.uint8)
bgc = 250, 85, 70
bg[:] = bgc
#bg[:] = 0
from skimage import draw
line = draw.line(0, 0, 10, 10)
def drawXyCycle(img, yx=(200, 300), r=8, color=255):
cir = draw.circle(*yx, r)
cir = list(cir)
cir[0] = cir[0].clip(0, img.shape[0] - 1)
cir[1] = cir[1].clip(0, img.shape[1] - 1)
batchPerGpu = 32
channl = 17
gpun = th.cuda.device_count()
batch = batchPerGpu * gpun
cyc_r = intround((25) / np.array([353, 257]).mean() * hw.mean())
pgt = hw // 2
pshift = (hw * .13).intround()
pshift = Vector([0, 0])
pre = pgt + pshift
coreShape = Vector((
cyc_r * 2,
cyc_r * 2,
))
feat = np.zeros([batch, channl] + list(hw))
raw_cycle_tmpl = getWeightCore(300, seta=.25)
# raw_cycle_tmpl = raw_cycle_tmpl > raw_cycle_tmpl[150,0]
core = resize(raw_cycle_tmpl, coreShape)
msegt = feat.copy()
msegt[..., (pgt - coreShape // 2).x:(pgt + coreShape // 2).x,
(pgt - coreShape // 2).y:(pgt + coreShape // 2).y, ] = core
feat[..., (pre - coreShape // 2).x:(pre + coreShape // 2).x,
(pre - coreShape // 2).y:(pre + coreShape // 2).y, ] = core
msegt = torch.tensor(msegt)
feat = torch.tensor(feat)
# show(core, feat)
from boxx import makedirs, np, npa, imread, pathjoin, glob, os, resize, p, uint8
ignoreWarning()
celebA_path = os.path.expanduser('~/dataset/celeba')
dataset = pathjoin(celebA_path, 'eyeglasses_stgan_dataset')
st_gan_dataset = pathjoin(dataset, 'tf_st_gan_dataset')
psa = sorted(glob(pathjoin(dataset, 'trainA/*')))[0::2]
psb = sorted(glob(pathjoin(dataset, 'trainB/*')))[1::2]
lena = len(psa)
lenb = len(psb)
attribute = np.zeros((lena + lenb, 40), np.bool)
attribute[-lenb:, 15] = True
img = imread(psa[0])
shape = img.shape
makedirs(p / st_gan_dataset)
imgns = psa + psb
if shape[:2] != (
218,
178,
):
def f(imgn):
return uint8(resize(imread(imgn), (
def visBboxList(rgb,
bboxList,
path='/tmp/visCanvas/tmp.pdf',
classNames=None,
box_alpha=0.8,
show_class=True,
thresh=0.1,
show_mask=None,
pltshow=False):
if classNames is None:
classNames = {i: "%d-class" % i for i in range(1111)}
classn = len(classNames)
dataset = dicto(classes=classNames)
cls_segms = [[] for _ in range(classn)]
cls_boxes = [np.zeros((0, 5), np.float32) for _ in range(classn)]
#cls_segms = None
#cls_boxes = [np.zeros((0,5), np.float32) for _ in range(classn)]
bboxnps = bboxList.bbox
extraFields = {
k: v.cpu().numpy()
for k, v in bboxList.extra_fields.items()
}
if show_mask is None:
if 'mask' in extraFields:
show_mask = True
for ind, bboxnp in enumerate(bboxnps):
# g()
other = dicto({k: v[ind] for k, v in extraFields.items()})
if other.scores < thresh:
continue
c = other.labels
if show_mask:
rle = mask2rle(other.mask)
cls_segms[c].append(rle)
if bboxList.mode == 'xyxy':
cls_boxes[c] = np.append(cls_boxes[c],
[list(bboxnp) + [other.scores]], 0)
cls_keyps = None
if not show_mask:
cls_segms = None
# g()
outputDir, name = dirname(path), filename(path)
vis_one_image(
rgb, # BGR -> RGB for visualization
name,
outputDir,
cls_boxes,
cls_segms,
cls_keyps,
dataset=dataset,
box_alpha=0.8,
show_class=True,
thresh=thresh,
kp_thresh=2,
pltshow=pltshow,
)