YOLOv1 with Tensorflow 2

Paper :- You Only Look Once: Unified, Real-Time Object Detection

DagsHub Repository

For ease of implementation, i have not implemented exactly the same as paper.The things presented below are implemented differently from the paper.

• Backbone network. (Used Xception instead of network mentioned in the paper.)

• Learning rate schedule (Used tf.keras.optimizers.schedules.ExponentialDecay)

• Data augmentations

• Hyper parameters

• And so on . . .

Preview

Tensorboard

Inference Visualization

Build Environment with Docker

Build Docker Image

$ docker build -t ${NAME}:${TAG} .

Create a Container

$ docker run -d -it --gpus all --shm-size=${PROPER_VALUE} ${NAME}:${TAG} /bin/bash

Training Dataset: Pascal VOC Dataset (VOC2008 | VOC2012)

Pascal VOC Dataset with TFDS

Number of Images

	Train	Validation	Test
Pascal VOC 2008	2501	2510	4952 (Used Validation)
Pascal VOC 2012	5717	5823	10991 (No labels)

• Training Set: VOC2008 trainval + VOC2012 trainval (Total: 16551)
• Validation Set: VOC2008 test (Total: 4952)

Pretrained PB File

Trained with default values of this repo. (Total Epoch: 105)

• Download pb file: <Link>

pb file is uploaded as tar.gz. So, you have to decompress this file like below.

$ tar -zxvf yolo_voc_448x448.tar.gz

If you want to inference with this pb file, infer to inference_tutorial.ipynb

Performance

Evaluation with VOC2008 test

class name	AP
dog	0.7464
pottedplant	0.2250
car	0.5021
person	0.4482
tvmonitor	0.5213
diningtable	0.4564
bicycle	0.5927
chair	0.2041
motorbike	0.5595
sofa	0.4801
bus	0.6215
boat	0.3274
horse	0.7049
aeroplane	0.5872
sheep	0.4223
bottle	0.1312
train	0.7917
cat	0.8044
bird	0.4824
cow	0.4548
mAP	0.5032

Inference Speed

• GPU(GeForce RTX 3090): About 8 FPS
• CPU(AMD Ryzen 5 5600X 6-Core Processor): About 4 FPS

Training Script

Path: ./voc_scripts/train_voc.py

$ python train_voc.py

Options

Default option values are ./configs/configs.py.If the options are given, the default config values are overridden.

• --epochs: Number of training epochs
• --init_lr: Initial learning rate
• --lr_decay_rate: Learning rate decay rate
• --lr_decay_steps: Learning rate decay steps
• --batch_size: Training batch size
• --val_step: Validation interval during training
• --tb_img_max_outputs : Number of visualized prediction images in tensorboard
• --train_ds_sample_ratio: Training dataset sampling ratio
• --val_ds_sample_ratio: Validation dataset sampling ratio

Evaluation Script

Path: ./voc_scripts/eval_voc.py

Evaluation pretrained model with VOC2008 Test Dataset

$ python eval_voc.py

Options

• --batch_size: Evaluation batch size (Default: batch_size of ./configs/configs.py.)
• --pb_dir: Save pb directory path (Default: ./ckpts/voc_ckpts/yolo_voc_448x448)

Citation

You Only Look Once: Unified, Real-Time Object Detection <arxiv link>

@misc{redmon2016look,
      title={You Only Look Once: Unified, Real-Time Object Detection}, 
      author={Joseph Redmon and Santosh Divvala and Ross Girshick and Ali Farhadi},
      year={2016},
      eprint={1506.02640},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}