pytorch实现yolov3(4) 非极大值抑制nms
在上一篇里咱们实现了forward函数.获得了prediction.此时预测出了特别多的box以及各类class probability,如今咱们要从中过滤出咱们最终的预测box.
理解了yolov3的输出的格式及每个位置的含义,并不难理解源码.我在阅读源码的过程当中主要的困难在于对pytorch不熟悉,因此在这篇文章里,关于其中涉及的一些pytorch中的函数的用法我都已经用加粗标示了而且给出了相应的连接,测试代码等.html
obj score threshold
咱们设置一个obj score thershold,超过这个值的才认为是有效的.ide
conf_mask = (prediction[:,:,4] > confidence).float().unsqueeze(2)
prediction = prediction*conf_mask
prediction是1*boxnum*boxattr
prediction[:,:,4]是1*boxnum 元素值为boxattr的index=4的那个值.函数
torch中的Tensor index和numpy是相似的,参看下列代码输出oop
import torch
x = torch.Tensor(1,3,10) # Create an un-initialized Tensor of size 2x3
print(x)
print(x.shape) # Print out the Tensor
y = x[:,:,4]
print(y)
print(y.shape)
z = x[:,:,4:6]
print(z)
print(z.shape)
print((y>0.5).float().unsqueeze(2))
#### 输出以下
tensor([[[2.5226e-18, 1.6898e-04, 1.0413e-11, 7.7198e-10, 1.0549e-08,
4.0516e-11, 1.0681e-05, 2.9575e-18, 6.7333e+22, 1.7591e+22],
[1.7184e+25, 4.3222e+27, 6.1972e-04, 7.2443e+22, 1.7728e+28,
7.0367e+22, 5.9018e-10, 2.6540e-09, 1.2972e-11, 5.3370e-08],
[2.7001e-06, 2.6801e-09, 4.1292e-05, 2.1511e+23, 3.2770e-09,
2.5125e-18, 7.7052e+31, 1.9447e+31, 5.0207e+28, 1.1492e-38]]])
torch.Size([1, 3, 10])
tensor([[1.0549e-08, 1.7728e+28, 3.2770e-09]])
torch.Size([1, 3])
tensor([[[1.0549e-08, 4.0516e-11],
[1.7728e+28, 7.0367e+22],
[3.2770e-09, 2.5125e-18]]])
torch.Size([1, 3, 2])
tensor([[[0.],
[0.],
[0.]]])
Squeeze and unsqueeze 下降维度,升高维度.测试
t = torch.ones(2,1,2,1) # Size 2x1x2x1 r = torch.squeeze(t) # Size 2x2 r = torch.squeeze(t, 1) # Squeeze dimension 1: Size 2x2x1 # Un-squeeze a dimension x = torch.Tensor([1, 2, 3]) r = torch.unsqueeze(x, 0) # Size: 1x3 表示在第0个维度添加1维 r = torch.unsqueeze(x, 1) # Size: 3x1 表示在第1个维度添加1维
这样prediction中objscore<threshold的已经变成了0.ui
nms
tensor.new() 建立一个和原有tensor的dtype一致的新tensor https://stackoverflow.com/questions/49263588/pytorch-beginner-tensor-new-method.net
#获得box坐标(top-left corner x, top-left corner y, right-bottom corner x, right-bottom corner y)
box_corner = prediction.new(prediction.shape)
box_corner[:,:,0] = (prediction[:,:,0] - prediction[:,:,2]/2)
box_corner[:,:,1] = (prediction[:,:,1] - prediction[:,:,3]/2)
box_corner[:,:,2] = (prediction[:,:,0] + prediction[:,:,2]/2)
box_corner[:,:,3] = (prediction[:,:,1] + prediction[:,:,3]/2)
prediction[:,:,:4] = box_corner[:,:,:4]
原始的prediction中boxattr存放的是x,y,w,h,...,不方便咱们处理,咱们将其转换成(top-left corner x, top-left corner y, right-bottom corner x, right-bottom corner y)scala
接下来咱们挨个处理每一张图片对应的feature map.code
batch_size = prediction.size(0)
write = False
for ind in range(batch_size):
#image_pred.shape=boxnum\*boxattr
image_pred = prediction[ind] #image Tensor box_num*box_attr
#confidence threshholding
#NMS
#返回每一行的最大值,及最大值所在的列.
max_conf, max_conf_score = torch.max(image_pred[:,5:5+ num_classes], 1)
#升级成和image_pred一样的维度
max_conf = max_conf.float().unsqueeze(1)
max_conf_score = max_conf_score.float().unsqueeze(1)
seq = (image_pred[:,:5], max_conf, max_conf_score)
#沿着列的方向拼接. 如今image_pred变成boxnum\*7
image_pred = torch.cat(seq, 1)
这里涉及到torch.max的用法,参见https://blog.csdn.net/Z_lbj/article/details/79766690
torch.max(input, dim, keepdim=False, out=None) -> (Tensor, LongTensor)
按维度dim 返回最大值.能够这么记忆,沿着第dim维度比较.torch.max(0)即沿着行的方向比较,即获得每列的最大值.
假设input是二维矩阵,即行*列,行是第0维,列是第一维.orm
- torch.max(a,0) 返回每一列中最大值的那个元素，且返回索引（返回最大元素在这一列的行索引）
- torch.max(a,1) 返回每一行中最大值的那个元素，且返回其索引（返回最大元素在这一行的列索引）
c=torch.Tensor([[1,2,3],[6,5,4]])
print(c)
a,b=torch.max(c,1)
print(a)
print(b)
##输出以下:
tensor([[1., 2., 3.],
[6., 5., 4.]])
tensor([3., 6.])
tensor([2, 0])
torch.cat用法,参见https://pytorch.org/docs/stable/torch.html
torch.cat(tensors, dim=0, out=None) → Tensor
>>> x = torch.randn(2, 3)
>>> x
tensor([[ 0.6580, -1.0969, -0.4614],
[-0.1034, -0.5790, 0.1497]])
>>> torch.cat((x, x, x), 0)
tensor([[ 0.6580, -1.0969, -0.4614],
[-0.1034, -0.5790, 0.1497],
[ 0.6580, -1.0969, -0.4614],
[-0.1034, -0.5790, 0.1497],
[ 0.6580, -1.0969, -0.4614],
[-0.1034, -0.5790, 0.1497]])
>>> torch.cat((x, x, x), 1)
tensor([[ 0.6580, -1.0969, -0.4614, 0.6580, -1.0969, -0.4614, 0.6580,
-1.0969, -0.4614],
[-0.1034, -0.5790, 0.1497, -0.1034, -0.5790, 0.1497, -0.1034,
-0.5790, 0.1497]])
接下来咱们只处理obj_score非0的数据(obj_score<obj_threshold转变为0)
non_zero_ind = (torch.nonzero(image_pred[:,4]))
try:
image_pred_ = image_pred[non_zero_ind.squeeze(),:].view(-1,7)
except:
continue
#For PyTorch 0.4 compatibility
#Since the above code with not raise exception for no detection
#as scalars are supported in PyTorch 0.4
if image_pred_.shape[0] == 0:
continue
ok,接下来咱们对每一种class作nms.
首先取到咱们有哪些类别
#Get the various classes detected in the image
img_classes = unique(image_pred_[:,-1]) # -1 index holds the class index
而后依次对每一种类别作处理
for cls in img_classes:
#perform NMS
#get the detections with one particular class
#取出当前class为当前class且class prob!=0的行
cls_mask = image_pred_*(image_pred_[:,-1] == cls).float().unsqueeze(1)
class_mask_ind = torch.nonzero(cls_mask[:,-2]).squeeze()
image_pred_class = image_pred_[class_mask_ind].view(-1,7)
#sort the detections such that the entry with the maximum objectness
#confidence is at the top
#按照obj score从高到低作排序
conf_sort_index = torch.sort(image_pred_class[:,4], descending = True )[1]
image_pred_class = image_pred_class[conf_sort_index]
idx = image_pred_class.size(0) #Number of detections
for i in range(idx):
#Get the IOUs of all boxes that come after the one we are looking at
#in the loop
try:
#计算第i个和其后每一行的的iou
ious = bbox_iou(image_pred_class[i].unsqueeze(0), image_pred_class[i+1:])
except ValueError:
break
except IndexError:
break
#Zero out all the detections that have IoU > treshhold
#把与第i行iou>nms_conf的认为是同一个目标的box,将其转成0
iou_mask = (ious < nms_conf).float().unsqueeze(1)
image_pred_class[i+1:] *= iou_mask
#把iou>nms_conf的移除掉
non_zero_ind = torch.nonzero(image_pred_class[:,4]).squeeze()
image_pred_class = image_pred_class[non_zero_ind].view(-1,7)
batch_ind = image_pred_class.new(image_pred_class.size(0), 1).fill_(ind) #Repeat the batch_id for as many detections of the class cls in the image
seq = batch_ind, image_pred_class
其中计算iou的代码以下,很少解释了.iou=交叠面积/总面积
def bbox_iou(box1, box2):
"""
Returns the IoU of two bounding boxes
"""
#Get the coordinates of bounding boxes
b1_x1, b1_y1, b1_x2, b1_y2 = box1[:,0], box1[:,1], box1[:,2], box1[:,3]
b2_x1, b2_y1, b2_x2, b2_y2 = box2[:,0], box2[:,1], box2[:,2], box2[:,3]
#get the corrdinates of the intersection rectangle
inter_rect_x1 = torch.max(b1_x1, b2_x1)
inter_rect_y1 = torch.max(b1_y1, b2_y1)
inter_rect_x2 = torch.min(b1_x2, b2_x2)
inter_rect_y2 = torch.min(b1_y2, b2_y2)
#Intersection area
inter_area = torch.clamp(inter_rect_x2 - inter_rect_x1 + 1, min=0) * torch.clamp(inter_rect_y2 - inter_rect_y1 + 1, min=0)
#Union Area
b1_area = (b1_x2 - b1_x1 + 1)*(b1_y2 - b1_y1 + 1)
b2_area = (b2_x2 - b2_x1 + 1)*(b2_y2 - b2_y1 + 1)
iou = inter_area / (b1_area + b2_area - inter_area)
return iou
关于nms能够看下https://blog.csdn.net/shuzfan/article/details/52711706
tensor index操做用法以下:
image_pred_ = torch.Tensor([[1,2,3,4,9],[5,6,7,8,9]])
#print(image_pred_[:,-1] == 9)
has_9 = (image_pred_[:,-1] == 9)
print(has_9)
###执行顺序是(image_pred_[:,-1] == 9).float().unsqueeze(1) 再作tensor乘法
cls_mask = image_pred_*(image_pred_[:,-1] == 9).float().unsqueeze(1)
print(cls_mask)
class_mask_ind = torch.nonzero(cls_mask[:,-2]).squeeze()
image_pred_class = image_pred_[class_mask_ind]
输出以下:
tensor([1, 1], dtype=torch.uint8)
tensor([[1., 2., 3., 4., 9.],
[5., 6., 7., 8., 9.]])
torch.sort用法以下:
d=torch.Tensor([[1,2,3],[6,5,4]]) e=d[:,2] print(e) print(torch.sort(e)) 输出 tensor([3., 4.]) torch.return_types.sort( values=tensor([3., 4.]), indices=tensor([0, 1]))
总结一下咱们作nms的流程
每个image,会预测出N个detetction信息,包括4+1+C(4个坐标信息,1个obj score以及C个class probability)
- 首先过滤掉obj_score < confidence的行
- 每一行只取class probability最高的做为预测出来的类别
- 将全部的预测按照obj_score从大到小排序
- 循环每一种类别,开始作nms
- 比较第一个box与其后全部box的iou,删除iou>threshold的box,即剔除全部类似box
- 比较下一个box与其后全部box的iou,删除全部与该box类似的box
- 不断重复上述过程,直至再也不有类似box
- 至此,实现了当前处理的类别的多个box均是独一无二的box.
write_results最终的返回值是一个n*8的tensor,其中8是(batch_index,4个坐标,1个objscore,1个class prob,一个class index)
def write_results(prediction, confidence, num_classes, nms_conf = 0.4):
print("prediction.shape=",prediction.shape)
#将obj_score < confidence的行置为0
conf_mask = (prediction[:,:,4] > confidence).float().unsqueeze(2)
prediction = prediction*conf_mask
#获得box坐标(top-left corner x, top-left corner y, right-bottom corner x, right-bottom corner y)
box_corner = prediction.new(prediction.shape)
box_corner[:,:,0] = (prediction[:,:,0] - prediction[:,:,2]/2)
box_corner[:,:,1] = (prediction[:,:,1] - prediction[:,:,3]/2)
box_corner[:,:,2] = (prediction[:,:,0] + prediction[:,:,2]/2)
box_corner[:,:,3] = (prediction[:,:,1] + prediction[:,:,3]/2)
#修改prediction第三个维度的前四列
prediction[:,:,:4] = box_corner[:,:,:4]
batch_size = prediction.size(0)
write = False
for ind in range(batch_size):
#image_pred.shape=boxnum\*boxattr
image_pred = prediction[ind] #image Tensor
#confidence threshholding
#NMS
##取出每一行的class score最大的一个
max_conf_score,max_conf = torch.max(image_pred[:,5:5+ num_classes], 1)
max_conf = max_conf.float().unsqueeze(1)
max_conf_score = max_conf_score.float().unsqueeze(1)
seq = (image_pred[:,:5], max_conf_score, max_conf)
image_pred = torch.cat(seq, 1) #如今变成7列,分别为左上角x,左上角y,右下角x,右下角y,obj score,最大probabilty,相应的class index
print(image_pred.shape)
non_zero_ind = (torch.nonzero(image_pred[:,4]))
try:
image_pred_ = image_pred[non_zero_ind.squeeze(),:].view(-1,7)
except:
continue
#For PyTorch 0.4 compatibility
#Since the above code with not raise exception for no detection
#as scalars are supported in PyTorch 0.4
if image_pred_.shape[0] == 0:
continue
#Get the various classes detected in the image
img_classes = unique(image_pred_[:,-1]) # -1 index holds the class index
for cls in img_classes:
#perform NMS
#get the detections with one particular class
#取出当前class为当前class且class prob!=0的行
cls_mask = image_pred_*(image_pred_[:,-1] == cls).float().unsqueeze(1)
class_mask_ind = torch.nonzero(cls_mask[:,-2]).squeeze()
image_pred_class = image_pred_[class_mask_ind].view(-1,7)
#sort the detections such that the entry with the maximum objectness
#confidence is at the top
#按照obj score从高到低作排序
conf_sort_index = torch.sort(image_pred_class[:,4], descending = True )[1]
image_pred_class = image_pred_class[conf_sort_index]
idx = image_pred_class.size(0) #Number of detections
for i in range(idx):
#Get the IOUs of all boxes that come after the one we are looking at
#in the loop
try:
#计算第i个和其后每一行的的iou
ious = bbox_iou(image_pred_class[i].unsqueeze(0), image_pred_class[i+1:])
except ValueError:
break
except IndexError:
break
#Zero out all the detections that have IoU > treshhold
#把与第i行iou>nms_conf的认为是同一个目标的box,将其转成0
iou_mask = (ious < nms_conf).float().unsqueeze(1)
image_pred_class[i+1:] *= iou_mask
#把iou>nms_conf的移除掉
non_zero_ind = torch.nonzero(image_pred_class[:,4]).squeeze()
image_pred_class = image_pred_class[non_zero_ind].view(-1,7)
batch_ind = image_pred_class.new(image_pred_class.size(0), 1).fill_(ind) #Repeat the batch_id for as many detections of the class cls in the image
seq = batch_ind, image_pred_class
if not write:
output = torch.cat(seq,1) #沿着列方向,shape 1*8
write = True
else:
out = torch.cat(seq,1)
output = torch.cat((output,out)) #沿着行方向 shape n*8
try:
return output
except:
return 0
相关文章
- 1. NMS(非极大值抑制)
- 2. NMS--非极大值抑制
- 3. 非极大值抑制NMS
- 4. 非极大值抑制（NMS）
- 5. 非极大值抑制 nms
- 6. NMS——非极大值抑制
- 7. 非极大值抑制（NMS）详解
- 8. NMS（non maximum suppression）非极大值抑制
- 9. 非极大值抑制（NMS）讲解
- 10. 非极大值抑制（Non-Maximum Suppression，NMS）
- 更多相关文章...
- • 现实生活中的 XML - XML 教程
- • Hibernate实现增删改查 - Hibernate教程
- • ☆基于Java Instrument的Agent实现
- • Spring Cloud 微服务实战(三) - 服务注册与发现
相关标签/搜索
每日一句
-
每一个你不满意的现在，都有一个你没有努力的曾经。
欢迎关注本站公众号,获取更多信息
相关文章
- 1. NMS(非极大值抑制)
- 2. NMS--非极大值抑制
- 3. 非极大值抑制NMS
- 4. 非极大值抑制（NMS）
- 5. 非极大值抑制 nms
- 6. NMS——非极大值抑制
- 7. 非极大值抑制（NMS）详解
- 8. NMS（non maximum suppression）非极大值抑制
- 9. 非极大值抑制（NMS）讲解
- 10. 非极大值抑制（Non-Maximum Suppression，NMS）