Tools：IOU、NMS、RAdam、one-hot

Tools：IOU、NMS、RAdam、one-hot

▶IOU

def iou(box, boxes, isMin = False):
	area = (box[2] - box[0]) * (box[3] - box[1])	 # 第一个框和剩下 所有的框比
	areas = (boxes[:, 2] - boxes[:, 0]) * (boxes[:, 3] - boxes[:, 1])
	x1 = np.maximum(box[0], boxes[:, 0])	 # 交集坐标
	y1 = np.maximum(box[1], boxes[:, 1])
	x2 = np.minimum(box[2], boxes[:, 2])
	y2 = np.minimum(box[3], boxes[:, 3])
	w = np.maximum(0, x2 - x1)				# 判断交集，无交集时宽高为0
	h = np.maximum(0, y2 - y1)
	inter = w * h							# 交集面积

	if isMin:								# 是否大框套小框
		ovr = np.true_divide(inter, np.minimum(area, areas))
	else:
		ovr = np.true_divide(inter, (areas + area - inter))
	return ovr

▶NMS

def nms(boxes, threth = 0.3, isMin = False):
	"""
	防止传入空boxes:
	:param boxes:[x1, y1, x2, y2, c]
	:return:r_boxes
	"""
	r_boxes = [] 					# 存放最后需要留下并返回给调用方的框
	if boxes.shape[0] == 0:			# 如果传入空，则返回空
		return np.array([])
	_boxes = boxes[(-boxes[:, 0]).argsort()]	 # 传非空，排序
	# print('_boxes:', _boxes)
	while _boxes.shape[0] > 1:  	# 比较后删除无效的框，直到剩下最后一个框，只剩一个时需保留，且无法计算IOU
		a_box = _boxes[0]			# 取第一个框
		b_boxes = _boxes[1:]		# 取剩下的框
		r_boxes.append(a_box)		# 第一个框肯定要留，所以添加到返回的列表中
		index = np.where(iou(a_box[1:5], b_boxes[:, 1:5], isMin) < threth)	 # 返回交并比小于threth的数据的索引
		# break
		_boxes = b_boxes[index]		# 到此，排除了与第一个框交并比较大的，
		# 然后将剩下的框赋给_boxes，接下来取出剩下的第一个与其他比，如此循环
	if _boxes.shape[0] > 0:
		r_boxes.append(_boxes[0])
	r_boxes = np.stack(r_boxes)
	return r_boxes

▶one-hot（Pytorch提供了one-hot函数可调，torch.nn.functional.one_hot( ) ）

def one_hot1(x, num_class=None):			# 一维
	ohx = np.zeros(num_class)
	ohx[int(x)] = 1
	return ohx

def one_hot2(x, num_class=None):			# 二维
	if not num_class:
		num_class = numpy.max(x) + 1
	ohx = numpy.zeros((int(x), num_class))
	ohx[range(int(x)), int(x)] = 1
	return ohx[0]
	
def one_hot3(x, num_class=None):		# 二维
    return torch.zeros((len(x), num_class)).scatter(1, x, 1)

▶RAdam

class RAdam(Optimizer):

	def __init__(self, params, lr=1e-3, betas=(0.9, 0.999), eps=1e-8, weight_decay=0, degenerated_to_sgd=True):
		if not 0.0 <= lr:
			raise ValueError("Invalid learning rate: {}".format(lr))
		if not 0.0 <= eps:
			raise ValueError("Invalid epsilon value: {}".format(eps))
		if not 0.0 <= betas[0] < 1.0:
			raise ValueError("Invalid beta parameter at index 0: {}".format(betas[0]))
		if not 0.0 <= betas[1] < 1.0:
			raise ValueError("Invalid beta parameter at index 1: {}".format(betas[1]))

		self.degenerated_to_sgd = degenerated_to_sgd
		if isinstance(params, (list, tuple)) and len(params) > 0 and isinstance(params[0], dict):
			for param in params:
				if 'betas' in param and (param['betas'][0] != betas[0] or param['betas'][1] != betas[1]):
					param['buffer'] = [[None, None, None] for _ in range(10)]
		defaults = dict(lr=lr, betas=betas, eps=eps, weight_decay=weight_decay,
						buffer=[[None, None, None] for _ in range(10)])
		super(RAdam, self).__init__(params, defaults)

	def __setstate__(self, state):
		super(RAdam, self).__setstate__(state)

	def step(self, closure=None):

		loss = None
		if closure is not None:
			loss = closure()

		for group in self.param_groups:

			for p in group['params']:
				if p.grad is None:
					continue
				grad = p.grad.data.float()
				if grad.is_sparse:
					raise RuntimeError('RAdam does not support sparse gradients')

				p_data_fp32 = p.data.float()

				state = self.state[p]

				if len(state) == 0:
					state['step'] = 0
					state['exp_avg'] = torch.zeros_like(p_data_fp32)
					state['exp_avg_sq'] = torch.zeros_like(p_data_fp32)
				else:
					state['exp_avg'] = state['exp_avg'].type_as(p_data_fp32)
					state['exp_avg_sq'] = state['exp_avg_sq'].type_as(p_data_fp32)

				exp_avg, exp_avg_sq = state['exp_avg'], state['exp_avg_sq']
				beta1, beta2 = group['betas']

				exp_avg_sq.mul_(beta2).addcmul_(1 - beta2, grad, grad)
				exp_avg.mul_(beta1).add_(1 - beta1, grad)

				state['step'] += 1
				buffered = group['buffer'][int(state['step'] % 10)]
				if state['step'] == buffered[0]:
					N_sma, step_size = buffered[1], buffered[2]
				else:
					buffered[0] = state['step']
					beta2_t = beta2 ** state['step']
					N_sma_max = 2 / (1 - beta2) - 1
					N_sma = N_sma_max - 2 * state['step'] * beta2_t / (1 - beta2_t)
					buffered[1] = N_sma

					# more conservative since it's an approximated value
					if N_sma >= 5:
						step_size = math.sqrt(
							(1 - beta2_t) * (N_sma - 4) / (N_sma_max - 4) * (N_sma - 2) / N_sma * N_sma_max / (
										N_sma_max - 2)) / (1 - beta1 ** state['step'])
					elif self.degenerated_to_sgd:
						step_size = 1.0 / (1 - beta1 ** state['step'])
					else:
						step_size = -1
					buffered[2] = step_size

				# more conservative since it's an approximated value
				if N_sma >= 5:
					if group['weight_decay'] != 0:
						p_data_fp32.add_(-group['weight_decay'] * group['lr'], p_data_fp32)
					denom = exp_avg_sq.sqrt().add_(group['eps'])
					p_data_fp32.addcdiv_(-step_size * group['lr'], exp_avg, denom)
					p.data.copy_(p_data_fp32)
				elif step_size > 0:
					if group['weight_decay'] != 0:
						p_data_fp32.add_(-group['weight_decay'] * group['lr'], p_data_fp32)
					p_data_fp32.add_(-step_size * group['lr'], exp_avg)
					p.data.copy_(p_data_fp32)

		return loss