数据增强之RandAugment

数据增强之RandAugment

paper: ​​code: ​​常见的图像识别任务中，增广的过程一般都是作为预处理阶段的任务之一。往往由于数据集过大而造成极大的计算损耗障碍。此外，由于所处的阶段不同，这些方法无法像模型算法一样随意调整正则化强度（尽管数据增广的效果直接取决于模型和数据集的大小）。

传统自动数据增广的策略通常是基于小数据集在轻量模型上训练后再应用于更大的模型。这就在策略上造成了一定的限制约束。本文则解决了这两大限制。RandAugment可以将数据增广所产生的增量样本空间大大缩小，从而使其可与模型训练过程捆绑在一起完成，避免将其作为独立的预处理任务来完成。此外，本文设置了增广强度的正则化参数，可以根据不同的模型和数据集大小进行调整。RandAugment方法可以作为外置工具作用于不同的图像处理任务、数据集工作中。

在CIFAR-10/100、SVHN和ImageNet数据集上能持平甚至优于先前的自动数据增广方法性能。在ImageNet数据集上，Baseline采用EfficientNet-B7结构的精度为84%，而AutoAugment+Baseline的精度为84.4%，本文的RandAugment+Baseline则达到了85.0%的准确率，分别提升了1和0.6个百分点。在目标检测方面，Baseline采用ResNet结构，添加RandAugment的效果较Baseline和其他增广方法提高了1.0～1.3个百分点。在COCO数据集上的表现也有0~0.3%mAP的效果提升。最后，由于本文超参数的可解释性，RandAugment可以用来研究数据作用与模型、数据集大小之间的关系。

RandAugment

考虑到以往数据增强方法都包含30多个参数，团队也将关注点转移到了如何大幅减少数据增强的参数空间。

为了减少参数空间的同时保持数据(图像)的多样性，研究人员用无参数过程替代了学习的策略和概率。

这些策略和概率适用于每次变换(transformation)，该过程始终选择均匀概率为1/k的变换。

也就是说，给定训练图像的N个变换，RandAugment就能表示KN个潜在策略。

最后，需要考虑到的一组参数是每个增强失真(augmentation distortion)的大小。

研究人员采用线性标度来表示每个转换的强度。简单来说，就是每次变换都在0到10的整数范围内，其中，10表示给定变换的最大范围。

并假设一个单一的全局失真M(global distortion M)可能就足以对所有转换进行参数化。

这样，生成的算法便包含两个参数N和M，还可以用两行Python代码简单表示：

可以使用标准方法高效地进行超参数优化，但是考虑到极小的搜索空间，研究人员发现朴素网格搜索(naive grid search)是非常有效的。

实验结果

代码

import cv2import numpy as npimport cv2## aug functionsdef identity_func(img): return imgdef autocontrast_func(img, cutoff=2): ''' same output as PIL.ImageOps.autocontrast ''' n_bins = 256 def tune_channel(ch): n = ch.size cut = cutoff * n // 100 if cut == 0: high, low = ch.max(), ch.min() else: hist = cv2.calcHist([ch], [0], None, [n_bins], [0, n_bins]) low = np.argwhere(np.cumsum(hist) > cut) low = 0 if low.shape[0] == 0 else low[0] high = np.argwhere(np.cumsum(hist[::-1]) > cut) high = n_bins - 1 if high.shape[0] == 0 else n_bins - 1 - high[0] if high <= low: table = np.arange(n_bins) else: scale = (n_bins - 1) / (high - low) offset = -low * scale table = np.arange(n_bins) * scale + offset table[table < 0] = 0 table[table > n_bins - 1] = n_bins - 1 table = table.clip(0, 255).astype(np.uint8) return table[ch] channels = [tune_channel(ch) for ch in cv2.split(img)] out = cv2.merge(channels) return outdef equalize_func(img): ''' same output as PIL.ImageOps.equalize PIL's implementation is different from cv2.equalize ''' n_bins = 256 def tune_channel(ch): hist = cv2.calcHist([ch], [0], None, [n_bins], [0, n_bins]) non_zero_hist = hist[hist != 0].reshape(-1) step = np.sum(non_zero_hist[:-1]) // (n_bins - 1) if step == 0: return ch n = np.empty_like(hist) n[0] = step // 2 n[1:] = hist[:-1] table = (np.cumsum(n) // step).clip(0, 255).astype(np.uint8) return table[ch] channels = [tune_channel(ch) for ch in cv2.split(img)] out = cv2.merge(channels) return outdef rotate_func(img, degree, fill=(0, 0, 0)): ''' like PIL, rotate by degree, not radians ''' H, W = img.shape[0], img.shape[1] center = W / 2, H / 2 M = cv2.getRotationMatrix2D(center, degree, 1) out = cv2.warpAffine(img, M, (W, H), borderValue=fill) return outdef solarize_func(img, thresh=128): ''' same output as PIL.ImageOps.posterize ''' table = np.array([el if el < thresh else 255 - el for el in range(256)]) table = table.clip(0, 255).astype(np.uint8) out = table[img] return outdef color_func(img, factor=5): ''' same output as PIL.ImageEnhance.Color ''' ## implementation according to PIL definition, quite slow # degenerate = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)[:, :, np.newaxis] # out = blend(degenerate, img, factor) # M = ( # np.eye(3) * factor # + np.float32([0.114, 0.587, 0.299]).reshape(3, 1) * (1. - factor) # )[np.newaxis, np.newaxis, :] M = ( np.float32([ [0.886, -0.114, -0.114], [-0.587, 0.413, -0.587], [-0.299, -0.299, 0.701]]) * factor + np.float32([[0.114], [0.587], [0.299]]) ) out = np.matmul(img, M).clip(0, 255).astype(np.uint8) return outdef contrast_func(img, factor=5): """ same output as PIL.ImageEnhance.Contrast """ mean = np.sum(np.mean(img, axis=(0, 1)) * np.array([0.114, 0.587, 0.299])) table = np.array([( el - mean) * factor + mean for el in range(256) ]).clip(0, 255).astype(np.uint8) out = table[img] return outdef brightness_func(img, factor=2): ''' same output as PIL.ImageEnhance.Contrast ''' table = (np.arange(256, dtype=np.float32) * factor).clip(0, 255).astype(np.uint8) out = table[img] return outdef sharpness_func(img, factor=2): ''' The differences the this result and PIL are all on the 4 boundaries, the center areas are same ''' kernel = np.ones((3, 3), dtype=np.float32) kernel[1][1] = 5 kernel /= 13 degenerate = cv2.filter2D(img, -1, kernel) if factor == 0.0: out = degenerate elif factor == 1.0: out = img else: out = img.astype(np.float32) degenerate = degenerate.astype(np.float32)[1:-1, 1:-1, :] out[1:-1, 1:-1, :] = degenerate + factor * (out[1:-1, 1:-1, :] - degenerate) out = out.astype(np.uint8) return outdef shear_x_func(img, factor, fill=(0, 0, 0)): H, W = img.shape[0], img.shape[1] M = np.float32([[1, factor, 0], [0, 1, 0]]) out = cv2.warpAffine(img, M, (W, H), borderValue=fill, flags=cv2.INTER_LINEAR).astype(np.uint8) return outdef translate_x_func(img, offset=10, fill=(0, 0, 0)): ''' same output as PIL.Image.transform ''' H, W = img.shape[0], img.shape[1] M = np.float32([[1, 0, -offset], [0, 1, 0]]) out = cv2.warpAffine(img, M, (W, H), borderValue=fill, flags=cv2.INTER_LINEAR).astype(np.uint8) return outdef translate_y_func(img, offset, fill=(0, 0, 0)): ''' same output as PIL.Image.transform ''' H, W = img.shape[0], img.shape[1] M = np.float32([[1, 0, 0], [0, 1, -offset]]) out = cv2.warpAffine(img, M, (W, H), borderValue=fill, flags=cv2.INTER_LINEAR).astype(np.uint8) return outdef posterize_func(img, bits): ''' same output as PIL.ImageOps.posterize ''' out = np.bitwise_and(img, np.uint8(255 << (8 - bits))) return outdef shear_y_func(img, factor, fill=(0, 0, 0)): H, W = img.shape[0], img.shape[1] M = np.float32([[1, 0, 0], [factor, 1, 0]]) out = cv2.warpAffine(img, M, (W, H), borderValue=fill, flags=cv2.INTER_LINEAR).astype(np.uint8) return outdef cutout_func(img, pad_size, replace=(0, 0, 0)): replace = np.array(replace, dtype=np.uint8) H, W = img.shape[0], img.shape[1] rh, rw = np.random.random(2) pad_size = pad_size // 2 ch, cw = int(rh * H), int(rw * W) x1, x2 = max(ch - pad_size, 0), min(ch + pad_size, H) y1, y2 = max(cw - pad_size, 0), min(cw + pad_size, W) out = img.copy() out[x1:x2, y1:y2, :] = replace return out### level to argsdef enhance_level_to_args(MAX_LEVEL): def level_to_args(level): return ((level / MAX_LEVEL) * 1.8 + 0.1,) return level_to_argsdef shear_level_to_args(MAX_LEVEL, replace_value): def level_to_args(level): level = (level / MAX_LEVEL) * 0.3 if np.random.random() > 0.5: level = -level return (level, replace_value) return level_to_argsdef translate_level_to_args(translate_const, MAX_LEVEL, replace_value): def level_to_args(level): level = (level / MAX_LEVEL) * float(translate_const) if np.random.random() > 0.5: level = -level return (level, replace_value) return level_to_argsdef cutout_level_to_args(cutout_const, MAX_LEVEL, replace_value): def level_to_args(level): level = int((level / MAX_LEVEL) * cutout_const) return (level, replace_value) return level_to_argsdef solarize_level_to_args(MAX_LEVEL): def level_to_args(level): level = int((level / MAX_LEVEL) * 256) return (level, ) return level_to_argsdef none_level_to_args(level): return ()def posterize_level_to_args(MAX_LEVEL): def level_to_args(level): level = int((level / MAX_LEVEL) * 4) return (level, ) return level_to_argsdef rotate_level_to_args(MAX_LEVEL, replace_value): def level_to_args(level): level = (level / MAX_LEVEL) * 30 if np.random.random() < 0.5: level = -level return (level, replace_value) return level_to_argsfunc_dict = { 'Identity': identity_func, 'AutoContrast': autocontrast_func, 'Equalize': equalize_func, # 'Rotate': rotate_func, 'Solarize': solarize_func, 'Color': color_func, 'Contrast': contrast_func, 'Brightness': brightness_func, 'Sharpness': sharpness_func, # 'ShearX': shear_x_func, # 'TranslateX': translate_x_func, # 'TranslateY': translate_y_func, 'Posterize': posterize_func, # 'ShearY': shear_y_func,}translate_const = 10MAX_LEVEL = 10replace_value = (128, 128, 128)arg_dict = { 'Identity': none_level_to_args, 'AutoContrast': none_level_to_args, 'Equalize': none_level_to_args, # 'Rotate': rotate_level_to_args(MAX_LEVEL, replace_value), 'Solarize': solarize_level_to_args(MAX_LEVEL), 'Color': enhance_level_to_args(MAX_LEVEL), 'Contrast': enhance_level_to_args(MAX_LEVEL), 'Brightness': enhance_level_to_args(MAX_LEVEL), 'Sharpness': enhance_level_to_args(MAX_LEVEL), # 'ShearX': shear_level_to_args(MAX_LEVEL, replace_value), # 'TranslateX': translate_level_to_args( # translate_const, MAX_LEVEL, replace_value # ), # 'TranslateY': translate_level_to_args( # translate_const, MAX_LEVEL, replace_value # ), 'Posterize': posterize_level_to_args(MAX_LEVEL), # 'ShearY': shear_level_to_args(MAX_LEVEL, replace_value),}class RandomAugment(object): def __init__(self, N=2, M=10): self.N = N self.M = M def get_random_ops(self): sampled_ops = np.random.choice(list(func_dict.keys()), self.N) return [(op, 1., self.M) for op in sampled_ops] def __call__(self, img): ops = self.get_random_ops() for name, prob, level in ops: if np.random.random() > prob: continue args = arg_dict[name](level) img = func_dict[name](img, *args) # img = cutout_func(img, 16, replace_value) return imgif __name__ == '__main__': import matplotlib.pyplot as plt a = RandomAugment() # img = np.random.randn(32, 32, 3) # a(img) imgPath = 'Subset1_img__3.png' img = cv2.imread(imgPath) img2 = a(img) plt.subplot(121) plt.imshow(img) plt.subplot(122) plt.imshow(img2) plt.show()

Reference

​​https://zhuanlan.zhihu.com/p/92751138​​

版权声明：本文内容由网络用户投稿，版权归原作者所有，本站不拥有其著作权，亦不承担相应法律责任。如果您发现本站中有涉嫌抄袭或描述失实的内容，请联系我们jiasou666@gmail.com 处理，核实后本网站将在24小时内删除侵权内容。