Hua
f5be7671bc
Split the single "3d" captcha type into three independent expert models: - 3d_text: 3D perspective text OCR (renamed from old "3d", CTC-based ThreeDCNN) - 3d_rotate: rotation angle regression (new RegressionCNN, circular loss) - 3d_slider: slider offset regression (new RegressionCNN, SmoothL1 loss) CAPTCHA_TYPES expanded from 3 to 5 classes. Classifier samples updated to 50000 (10000 per class). New generators, model, dataset, training utilities, and full pipeline/export/CLI support for all subtypes. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
14 lines
687 B
Python
14 lines
687 B
Python
"""
|
|
训练脚本包
|
|
|
|
- dataset.py: CRNNDataset / CaptchaDataset / RegressionDataset 通用数据集类
|
|
- train_utils.py: CTC 训练通用逻辑 (train_ctc_model)
|
|
- train_regression_utils.py: 回归训练通用逻辑 (train_regression_model)
|
|
- train_normal.py: 训练普通字符识别 (LiteCRNN - normal)
|
|
- train_math.py: 训练算式识别 (LiteCRNN - math)
|
|
- train_3d_text.py: 训练 3D 立体文字识别 (ThreeDCNN)
|
|
- train_3d_rotate.py: 训练 3D 旋转回归 (RegressionCNN)
|
|
- train_3d_slider.py: 训练 3D 滑块回归 (RegressionCNN)
|
|
- train_classifier.py: 训练调度分类器 (CaptchaClassifier)
|
|
"""
|