yolov3-tf2

yolov3-tf2 copied to clipboard

怎么提高识别率呢？

m480399775

yolov3 tiny - evaluation on pretrained weights gives lower accuracy than expected

1

According to https://pjreddie.com/darknet/yolo/ website accuracy on pre-trained weights using Coco dataset should be 0.331 mAP: Model | Train | Test | mAP | FLOPS | FPS | Cfg | Weights...

chenhayat

How to create tfrecord for coco dataset.

1

I am looking to train this on coco dataset but how to create tfrecord for coco dataset. No code is given for coco tfrecord creation only for VOC tfrecord creation...

NaveenVinayakS

Bumps [tensorflow](https://github.com/tensorflow/tensorflow) from 2.5.1 to 2.7.2. Release notes Sourced from tensorflow's releases. TensorFlow 2.7.2 Release 2.7.2 This releases introduces several vulnerability fixes: Fixes a code injection in saved_model_cli (CVE-2022-29216) Fixes...

dependabot[bot]

Bumps [tensorflow-gpu](https://github.com/tensorflow/tensorflow) from 2.5.1 to 2.7.2. Release notes Sourced from tensorflow-gpu's releases. TensorFlow 2.7.2 Release 2.7.2 This releases introduces several vulnerability fixes: Fixes a code injection in saved_model_cli (CVE-2022-29216) Fixes...

dependabot[bot]

python train.py Error during training？why？

1

(yolov3-tf2-cpu) (venv) M:\MachineLearning\yolov3-tf2>python train.py --dataset ./data/voc2007_train_stone.tfrecord --val_dataset ./data/voc2007_val_stone.tfrecord --classes ./data/stone.names --num_classes 1 --mode fit --transfer darknet --batch_size 4 --epochs 20 --weights ./checkpoints/yolov3.tf --weights_num_classes 80 2021-04-24 13:50:14.338255: I tensorflow/core/platform/cpu_feature_guard.cc:142] Your CPU...

jiangxinufo

ValueError: Shapes (18,) and (255,) are incompatible

5

python train.py --dataset ./data/train.tfrecord --val_dataset ./data/_val.tfrecord --classes ./data/coco2.names --num_classes 1 --mode fit --transfer darknet --batch_size 16 --epochs 3 --weights ./checkpoints/yolov3.tf --weights_num_classes 1 Not able to retrain the model with custom...

BalajiSri12

xy_loss的损失函数对应darknet 源码应该是　K.binary_crossentropy(true_xy - pred_xy)

1

xy_loss = obj_mask * box_loss_scale * \ tf.reduce_sum(tf.square(true_xy - pred_xy), axis=-1)

luogantt

change # TODO: use binary_crossentropy instead class_loss = obj_mask * sparse_categorical_crossentropy( true_class_idx, pred_class) to true_class_onehot = tf.one_hot(tf.cast(true_class_idx, tf.int64), depth=classes, axis=-1) true_class_binary = tf.reshape(true_class_onehot, (tf.shape(y_true)[0], grid_size,grid_size,tf.shape(y_true)[3],-1, 1)) pred_class_binary = tf.reshape(pred_class, (tf.shape(y_true)[0],...

ZXTFINAL

When I try to run export_tflite.py, it cannot work and shows: I0223 16:11:15.828265 8204 export_tflite.py:48] tflite model loaded Windows fatal exception: access violation Current thread 0x0000200c (most recent call first):...

jharden13HOU