fix: bump binary blink open value

This commit is contained in:
Prohurtz 2024-04-22 21:17:47 -05:00
parent 2243e63f3c
commit b11485f3fb
4 changed files with 80 additions and 70 deletions

View File

@ -965,7 +965,7 @@ def External_Run_AHSF(frame_gray):
# (pupil_rect_coarse[0], pupil_rect_coarse[1]),
# (
# pupil_rect_coarse[0] + pupil_rect_coarse[2],
# pupil_rect_coarse[1] + pupil_rect_coarse[3],
# pupil_rect_coarse[1] + pupil_rect_coarse[so 3],
# ),
# (0, 255, 0),
# 2,

View File

@ -29,7 +29,8 @@ LICENSE: Summer Software Distribution License 1.0
import numpy as np
def BLINK(self):
def BLINK(self):
if self.blink_clear == True:
self.max_ints = []
@ -39,14 +40,18 @@ def BLINK(self):
intensity = np.sum(self.current_image_gray_clean)
if self.calibration_frame_counter == 300:
self.filterlist = [] #clear filter
self.filterlist = [] # clear filter
if len(self.filterlist) < 300:
self.filterlist.append(intensity)
else:
self.filterlist.pop(0)
self.filterlist.append(intensity)
if intensity >= np.percentile(self.filterlist, 99) or intensity <= np.percentile(self.filterlist, 1) and len(self.max_ints) >= 1: # filter abnormally high values
try: # I don't want this here but I cant get python to stop crying when it's not
if (
intensity >= np.percentile(self.filterlist, 99)
or intensity <= np.percentile(self.filterlist, 1)
and len(self.max_ints) >= 1
): # filter abnormally high values
try: # I don't want this here but I cant get python to stop crying when it's not
intensity = min(self.max_ints)
except:
pass
@ -54,7 +59,7 @@ def BLINK(self):
self.frames = self.frames + 1
if intensity > self.max_int:
self.max_int = intensity
if self.frames > 300: #TODO: test this number more (make it a setting??)
if self.frames > 300: # TODO: test this number more (make it a setting??)
self.max_ints.append(self.max_int)
if intensity < self.min_int:
self.min_int = intensity
@ -63,9 +68,11 @@ def BLINK(self):
if intensity > min(self.max_ints):
blinkvalue = 0.0
else:
blinkvalue = 0.7
blinkvalue = 0.8
try:
return blinkvalue
except:
return 0.7
# print(self.blinkvalue, self.max_int, self.min_int, self.frames, intensity)
return 0.8
# print(self.blinkvalue, self.max_int, self.min_int, self.frames, intensity)

View File

@ -1,4 +1,4 @@
'''
"""
------------------------------------------------------------------------------------------------------
,@@@@@@
@ -25,7 +25,7 @@ Algorithm App Implementations By: PallasNeko, Prohurtz
Copyright (c) 2023 EyeTrackVR <3
LICENSE: GNU GPLv3
------------------------------------------------------------------------------------------------------
'''
"""
import sys
from typing import Tuple
import math
@ -36,6 +36,7 @@ import onnxruntime
from one_euro_filter import OneEuroFilter
from utils.misc_utils import FastMedian, resource_path
import os
os.environ["OMP_NUM_THREADS"] = "1"
# DADDY
# Please change the name of this script and the name of the method if you have something better.
@ -44,15 +45,16 @@ input_size = 192 # Do not change this number.
heatmap_size = 48 # Do not change this number.
kernel_size = 7
if platform.system() == "Darwin":
model_file = "EyeTrackApp/Models/daddy230210.onnx" # The model file name will be changed when performance stabilises. # funny MacOS files issues :P
model_file = "Models/daddy230210.onnx" # The model file name will be changed when performance stabilises. # funny MacOS files issues :P
else:
model_file = "Models/daddy230210.onnx" # The model file name will be changed when performance stabilises.
# SHA256 for model version verification
# daddy230210.onnx = 59e59aa2a21024884200dd3acbd5e6a2e8d7209c46555fbdc727d4fe3adb68d3
imshow_enable = False
save_video = False
save_filepath = 'output.mp4'
save_filepath = "output.mp4"
def get_max_preds(batch_heatmaps):
@ -63,18 +65,18 @@ def get_max_preds(batch_heatmaps):
heatmaps_reshaped = batch_heatmaps.reshape((batch_size, num_joints, -1))
idx = np.argmax(heatmaps_reshaped, 2)
maxvals = np.amax(heatmaps_reshaped, 2)
maxvals = maxvals.reshape((batch_size, num_joints, 1))
idx = idx.reshape((batch_size, num_joints, 1))
preds = np.tile(idx, (1, 1, 2)).astype(np.float32)
preds[:, :, 0] = (preds[:, :, 0]) % width
preds[:, :, 1] = np.floor((preds[:, :, 1]) / width)
pred_mask = np.tile(np.greater(maxvals, 0.0), (1, 1, 2))
pred_mask = pred_mask.astype(np.float32)
preds *= pred_mask
return preds, maxvals
@ -89,12 +91,11 @@ def taylor(hm, coord):
dx = 0.5 * (hm[py][px + 1] - hm[py][px - 1])
dy = 0.5 * (hm[py + 1][px] - hm[py - 1][px])
dxx = 0.25 * (hm[py][px + 2] - 2 * hm[py][px] + hm[py][px - 2])
dxy = 0.25 * (hm[py + 1][px + 1] - hm[py - 1][px + 1] - hm[py + 1][px - 1] \
+ hm[py - 1][px - 1])
dxy = 0.25 * (hm[py + 1][px + 1] - hm[py - 1][px + 1] - hm[py + 1][px - 1] + hm[py - 1][px - 1])
dyy = 0.25 * (hm[py + 2 * 1][px] - 2 * hm[py][px] + hm[py - 2 * 1][px])
derivative = np.matrix([[dx], [dy]])
hessian = np.matrix([[dxx, dxy], [dxy, dyy]])
if dxx * dyy - dxy ** 2 != 0:
if dxx * dyy - dxy**2 != 0:
hessianinv = hessian.I
offset = -hessianinv * derivative
offset = np.squeeze(np.array(offset.T), axis=0)
@ -113,9 +114,9 @@ def gaussian_blur(hm, kernel):
for j in range(num_joints):
origin_max = np.max(hm[i, j])
dr = np.zeros((height + 2 * border, width + 2 * border))
dr[border: -border, border: -border] = hm[i, j].copy()
dr[border:-border, border:-border] = hm[i, j].copy()
dr = cv2.GaussianBlur(dr, (kernel, kernel), 0)
hm[i, j] = dr[border: -border, border: -border].copy()
hm[i, j] = dr[border:-border, border:-border].copy()
hm[i, j] *= origin_max / np.max(hm[i, j])
return hm
@ -123,7 +124,7 @@ def gaussian_blur(hm, kernel):
def get_final_preds(hm, realsize):
# base:https://github.com/ilovepose/DarkPose
coords, maxvals = get_max_preds(hm)
# post-processing
hm = gaussian_blur(hm, kernel_size)
hm = np.maximum(hm, 1e-10)
@ -131,22 +132,22 @@ def get_final_preds(hm, realsize):
for n in range(coords.shape[0]):
for p in range(coords.shape[1]):
coords[n, p] = taylor(hm[n][p], coords[n][p])
preds = coords.copy()
preds = (preds / heatmap_size) * realsize # input_size
# Transform back
# for i in range(coords.shape[0]):
# preds[i] = transform_preds(
# coords[i], center[i], scale[i], [heatmap_width, heatmap_height]
# )
return preds, maxvals
def resize_with_pad(image: np.array,
new_shape: Tuple[int, int],
padding_color: Tuple[int] = (255, 255, 255)) -> np.array:
def resize_with_pad(
image: np.array, new_shape: Tuple[int, int], padding_color: Tuple[int] = (255, 255, 255)
) -> np.array:
"""
https://gist.github.com/IdeaKing/11cf5e146d23c5bb219ba3508cca89ec
Maintains aspect ratio and resizes with padding.
@ -178,29 +179,32 @@ class BEER(object):
self.p03_med = FastMedian(k=256)
self.prev_ear = 0.5
# todo https://peerj.com/articles/cs-943/
def ear(self, pred):
p15 = np.linalg.norm(pred[1]-pred[5])
p24 = np.linalg.norm(pred[2]-pred[4])
p03 = np.linalg.norm(pred[0]-pred[3])
self.p03_med+p03
if p03 > self.p03_med.median()*1.5:
p15 = np.linalg.norm(pred[1] - pred[5])
p24 = np.linalg.norm(pred[2] - pred[4])
p03 = np.linalg.norm(pred[0] - pred[3])
self.p03_med + p03
if p03 > self.p03_med.median() * 1.5:
return self.prev_ear
ear = (p15+p24)/(2*self.p03_med.median())
ear = (p15 + p24) / (2 * self.p03_med.median())
self.ear_minmax(ear)
norm_ear = self.ear_norm(ear)
self.prev_ear = norm_ear.copy()
return norm_ear
def ear_minmax(self,ear):
def ear_minmax(self, ear):
if ear < self.ear_min:
self.ear_min = ear.copy()
if ear > self.ear_max:
self.ear_max = ear.copy()
def ear_norm(self,ear):
return (ear-self.ear_min)/(self.ear_max-self.ear_min) # todo:It is better to add very small values to avoid zero division.
def ear_norm(self, ear):
return (ear - self.ear_min) / (
self.ear_max - self.ear_min
) # todo:It is better to add very small values to avoid zero division.
#
# loopnum = 0
@ -215,35 +219,33 @@ class DADDY_cls(object):
options.intra_op_num_threads = 1 # This number should be changed accordingly
options.execution_mode = onnxruntime.ExecutionMode.ORT_SEQUENTIAL
options.graph_optimization_level = onnxruntime.GraphOptimizationLevel.ORT_ENABLE_ALL
ort_session = onnxruntime.InferenceSession(resource_path(model_file), sess_options=options, providers=["CPUExecutionProvider"])
ort_session.set_providers(['CPUExecutionProvider']) # only cpu mode
ort_session = onnxruntime.InferenceSession(
resource_path(model_file), sess_options=options, providers=["CPUExecutionProvider"]
)
ort_session.set_providers(["CPUExecutionProvider"]) # only cpu mode
self.ort_session = ort_session
self.input_name = ort_session.get_inputs()[0].name
self.output_name = ort_session.get_outputs()[0].name
min_cutoff = 0.0004
beta = 0.9
input_point = np.zeros((11, 2)) # np.array([1, 1])
self.one_euro_filter = OneEuroFilter(
input_point,
min_cutoff=min_cutoff,
beta=beta
)
self.one_euro_filter = OneEuroFilter(input_point, min_cutoff=min_cutoff, beta=beta)
# self.ear_oef = OneEuroFilter(
# np.zeros(1),
# min_cutoff=min_cutoff,
# beta=beta
# ) # memo: Parameters need tuning
self.beer = BEER()
# filepath = 'test.mp4'
# codec = cv2.VideoWriter_fourcc(*"mp4v")
# video = cv2.VideoWriter(filepath, codec, 60.0, (200, 150), 0) # (60, 60)) # (150, 200))
# self.video = video
def open_video(self, video_path):
# Temporary implementation to run
cap = cv2.VideoCapture(video_path)
@ -251,7 +253,7 @@ class DADDY_cls(object):
raise IOError("Error opening video stream or file")
self.cap = cap
return True
def read_frame(self):
# Temporary implementation to run
if not self.cap.isOpened():
@ -263,26 +265,28 @@ class DADDY_cls(object):
self.current_image_gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
return True
return False
def single_run(self):
# Temporary implementation to run
# todo: If it's the left hand eye, flip the image left to right.
gray_frame = self.current_image_gray.copy()
# frame_resize=resize_with_pad(gray_frame,(input_size,input_size))
# or
frame_resize = cv2.resize(gray_frame, (input_size, input_size))
imgs = np.divide(frame_resize[np.newaxis, np.newaxis], 255, dtype=np.float32) # input/255.0
pred_heatmap = self.ort_session.run(None, {self.input_name: imgs})[0] # .reshape((-1, 2))
# if imshow_enable:
# heatmap = pred_heatmap.reshape((-1, heatmap_size, heatmap_size))
# for i in range(heatmap.shape[0]):
# cv2.imshow("heatmap_{}".format(i + 1), heatmap[i])
pred, max_val = get_final_preds(pred_heatmap, (self.current_image_gray.shape[1], self.current_image_gray.shape[0]))
pred, max_val = get_final_preds(
pred_heatmap, (self.current_image_gray.shape[1], self.current_image_gray.shape[0])
)
pred = pred.reshape((-1, 2))
# or
# pred, max_val = get_final_preds(pred_heatmap, input_size)
@ -295,7 +299,7 @@ class DADDY_cls(object):
pred = self.one_euro_filter(pred)
kps = pred.astype(np.int32)
# eyecenter = kps[:6].mean(axis=0).astype(int)
ear = self.beer.ear(pred)
# ear=self.ear_oef(ear[np.newaxis])#memo: Parameters need tuning
@ -316,7 +320,6 @@ class DADDY_cls(object):
# cv2.putText(self.current_image_gray, str(i), (kps[i, 0] - 10, kps[i, 1] - 10), cv2.FONT_HERSHEY_SIMPLEX, 0.35, (0, 0, 255), 1)
# cv2.putText(self.current_image_gray, "EAR: "+str(ear), (self.current_image_gray.shape[1]//10, self.current_image_gray.shape[0]//10), cv2.FONT_HERSHEY_SIMPLEX, 0.75, (255,0,0), 1)
# global loopnum
# if loopnum < 1350*2:
# # self.video.write(cv2.resize(gray_frame.copy(), (200, 150), None))
@ -327,19 +330,19 @@ class DADDY_cls(object):
# sys.exit()
# if w_video:
# video.release()
# kps[i, :] = (x, y)
# i == [0:6] = Inner and outer corners of eyes and eyelids
# i == [6] = pupil
# i == [7:] = iris
return pupil_center_x, pupil_center_y, ear
class External_Run_DADDY(object):
def __init__(self):
self.algo = DADDY_cls()
def run(self, current_image_gray):
self.algo.current_image_gray = current_image_gray
pupil_x, pupil_y, ear = self.algo.single_run()
@ -350,4 +353,4 @@ if __name__ == "__main__":
daddy = DADDY_cls()
daddy.open_video(video_path)
while daddy.read_frame():
_ = daddy.single_run()
_ = daddy.single_run()

View File

@ -213,7 +213,7 @@ cct = 300
def RANSAC3D(self, hsrac_en):
f = False
ranf = False
blink = 0.7
blink = 0.8
angle = 0
if hsrac_en: