Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Summary translation #126

Open
wants to merge 14 commits into
base: master
Choose a base branch
from
2 changes: 1 addition & 1 deletion facebookresearch_WSL-Images_resnext.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ layout: hub_detail
background-class: hub-background
body-class: hub
title: ResNext WSL
summary: ResNext models trained with billion scale weakly-supervised data.
summary: 10억 규모의 약한 지도(weakly-supervised) 데이터셋을 사용한 ResNext 모델.
category: researchers
image: wsl-image.png
author: Facebook AI
Expand Down
2 changes: 1 addition & 1 deletion facebookresearch_pytorch-gan-zoo_pgan.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ layout: hub_detail
background-class: hub-background
body-class: hub
title: Progressive Growing of GANs (PGAN)
summary: High-quality image generation of fashion, celebrity faces
summary: 패션, 연예인 얼굴의 고품질 이미지 생성
category: researchers
image: pganlogo.png
author: FAIR HDGAN
Expand Down
9 changes: 5 additions & 4 deletions facebookresearch_pytorchvideo_resnet.md
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@ background-class: hub-background
body-class: hub
category: researchers
title: 3D ResNet
summary: Resnet Style Video classification networks pretrained on the Kinetics 400 dataset
summary: Kinetics 400 데이터셋에서 사전 학습된 Resnet 스타일 비디오 분류 네트워크
image: slowfast.png
author: FAIR PyTorchVideo
tags: [vision]
Expand All @@ -18,7 +18,7 @@ demo-model-link: https://huggingface.co/spaces/pytorch/3D_ResNet

### 사용 예시

#### Imports
#### 불러오기

모델 불러오기:

Expand Down Expand Up @@ -47,7 +47,7 @@ from pytorchvideo.transforms import (
)
```

#### 환경설정
#### 셋업

모델을 평가 모드로 설정하고 원하는 디바이스 방식을 선택합니다.

Expand All @@ -58,7 +58,7 @@ model = model.eval()
model = model.to(device)
```

토치 허브 모델이 훈련된 Kinetics 400 데이터셋에 대해 ID에서의 레이블과 맞는 정보를 다운로드합니다. 이는 예측된 클래스 ID에서 카테고리 레이블 이름을 가져오는데 사용됩니다.
토치 허브 모델이 훈련된 Kinetics 400 데이터셋에 대해 ID에서의 레이블 매핑 정보를 다운로드합니다. 이는 예측된 클래스 ID에서 카테고리 레이블 이름을 가져오는데 사용됩니다.

```python
json_url = "https://dl.fbaipublicfiles.com/pyslowfast/dataset/class_names/kinetics_classnames.json"
Expand Down Expand Up @@ -159,6 +159,7 @@ print("Top 5 predicted labels: %s" % ", ".join(pred_class_names))

### 모델 설명
모델 아키텍처는 Kinetics 데이터셋의 8x8 설정을 사용하여 사전 훈련된 가중치가 있는 참고문헌 [1]을 기반으로 합니다.

| arch | depth | frame length x sample rate | top 1 | top 5 | Flops (G) | Params (M) |
| --------------- | ----------- | ----------- | ----------- | ----------- | ----------- | ----------- |
| Slow | R50 | 8x8 | 74.58 | 91.63 | 54.52 | 32.45 |
Expand Down
4 changes: 2 additions & 2 deletions facebookresearch_pytorchvideo_slowfast.md
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@ background-class: hub-background
body-class: hub
category: researchers
title: SlowFast
summary: SlowFast networks pretrained on the Kinetics 400 dataset
summary: Kinetics 400 데이터섯에서 사전 학습된 SlowFast 네트워크
image: slowfast.png
author: FAIR PyTorchVideo
tags: [vision]
Expand Down Expand Up @@ -78,7 +78,7 @@ for k, v in kinetics_classnames.items():
kinetics_id_to_classname[v] = str(k).replace('"', "")
```

#### 입력 변환에 대한 정의
#### 입력 형태에 대한 정의

```python
side_size = 256
Expand Down
4 changes: 2 additions & 2 deletions facebookresearch_pytorchvideo_x3d.md
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@ background-class: hub-background
body-class: hub
category: researchers
title: X3D
summary: X3D networks pretrained on the Kinetics 400 dataset
summary: Kinetics 400 데이터섯에서 사전 학습된 X3D 네트워크
image: x3d.png
author: FAIR PyTorchVideo
tags: [vision]
Expand All @@ -18,7 +18,7 @@ demo-model-link: https://huggingface.co/spaces/pytorch/X3D

### 사용 예시

#### Imports
#### 불러오기

모델 불러오기:

Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ layout: hub_detail
background-class: hub-background
body-class: hub
title: Semi-supervised and semi-weakly supervised ImageNet Models
summary: Billion scale semi-supervised learning for image classification 에서 제안된 ResNet, ResNext 모델
summary: Billion scale semi-supervised learning for image classification 논문에서 제안된 ResNet, ResNext 모델
category: researchers
image: ssl-image.png
author: Facebook AI
Expand Down
2 changes: 1 addition & 1 deletion huggingface_pytorch-transformers.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ layout: hub_detail
background-class: hub-background
body-class: hub
title: PyTorch-Transformers
summary: PyTorch implementations of popular NLP Transformers
summary: 널리 사용되는 NLP Transformers의 PyTorch 구현
category: researchers
image: huggingface-logo.png
author: HuggingFace Team
Expand Down
18 changes: 13 additions & 5 deletions hustvl_yolop.md
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@ background-class: hub-background
body-class: hub
category: researchers
title: YOLOP
summary: YOLOP pretrained on the BDD100K dataset
summary: BDD100K 데이터 세트에서 사전 훈련된 YOLOP
image: yolop.png
author: Hust Visual Learning Team
tags: [vision]
Expand Down Expand Up @@ -125,7 +125,15 @@ det_out, da_seg_out,ll_seg_out = model(img)

### 인용(Citation)

See for more detail in [github code](https://github.com/hustvl/YOLOP) and [arxiv paper](https://arxiv.org/abs/2108.11250).

본 논문과 코드가 여러분의 연구에 유용하다고 판단되면, GitHub star를 주는 것과 본 논문을 인용하는 것을 고려해 주세요:

본 [논문](https://arxiv.org/abs/2108.11250) 과 [코드](https://github.com/hustvl/YOLOP) 가 여러분의 연구에 유용하다고 판단되면, GitHub star를 주는 것과 본 논문을 인용하는 것을 고려해 주세요:

```BibTeX
@article{wu2022yolop,
title={Yolop: You only look once for panoptic driving perception},
author={Wu, Dong and Liao, Man-Wen and Zhang, Wei-Tian and Wang, Xing-Gang and Bai, Xiang and Cheng, Wen-Qing and Liu, Wen-Yu},
journal={Machine Intelligence Research},
pages={1--13},
year={2022},
publisher={Springer}
}
```
2 changes: 1 addition & 1 deletion intelisl_midas_v2.md
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@ background-class: hub-background
body-class: hub
category: researchers
title: MiDaS
summary: MiDaS models for computing relative depth from a single image.
summary: 단일 이미지에서 상대적인 깊이를 계산하기 위한 MiDaaS 모델.
image: intel-logo.png
author: Intel ISL
tags: [vision]
Expand Down
2 changes: 1 addition & 1 deletion mateuszbuda_brain-segmentation-pytorch_unet.md
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@ background-class: hub-background
body-class: hub
category: researchers
title: U-Net for brain MRI
summary: U-Net with batch normalization for biomedical image segmentation with pretrained weights for abnormality segmentation in brain MRI
summary: 뇌 MRI 이상 분할을 위해 사전 훈련된 가중치를 사용한 의료 이미지 분할을 위한 배치 정규화를 사용하는 U-Net
image: unet_tcga_cs_4944.png
author: mateuszbuda
tags: [vision]
Expand Down
2 changes: 1 addition & 1 deletion nicolalandro_ntsnet-cub200_ntsnet.md
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@ background-class: hub-background
body-class: hub
category: researchers
title: ntsnet
summary: classify birds using this fine-grained image classifier
summary: fine-grained 이미지 분류기를 사용한 새 분류
image: Cub200Dataset.png
author: Moreno Caraffini and Nicola Landro
tags: [vision]
Expand Down
2 changes: 1 addition & 1 deletion nvidia_deeplearningexamples_efficientnet.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ layout: hub_detail
background-class: hub-background
body-class: hub
title: EfficientNet
summary: EfficientNets are a family of image classification models, which achieve state-of-the-art accuracy, being an order-of-magnitude smaller and faster. Trained with mixed precision using Tensor Cores.
summary: EfficientNets는 최첨단 정확도를 달성하는 이미지 분류 모델 계열로 크기가 작고 빠릅니다. 텐서 코어를 사용하여 혼합 정밀도로 훈련되었습니다.
category: researchers
image: nvidia_logo.png
author: NVIDIA
Expand Down
2 changes: 1 addition & 1 deletion nvidia_deeplearningexamples_resnet50.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ layout: hub_detail
background-class: hub-background
body-class: hub
title: ResNet50
summary: ResNet50 model trained with mixed precision using Tensor Cores.
summary: 텐서 코어를 사용하여 혼합 정밀도로 훈련된 ResNet50 모델.
category: researchers
image: nvidia_logo.png
author: NVIDIA
Expand Down
2 changes: 1 addition & 1 deletion nvidia_deeplearningexamples_resnext.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ layout: hub_detail
background-class: hub-background
body-class: hub
title: ResNeXt101
summary: ResNet with bottleneck 3x3 Convolutions substituted by 3x3 Grouped Convolutions, trained with mixed precision using Tensor Cores.
summary: ResNet의 3x3 그룹 합성곱(Grouped Convolution) 계층을 병목 블록(Bottleneck Block) 내부의 3x3 합성곱 계층으로 대체한 모델.
category: researchers
image: nvidia_logo.png
author: NVIDIA
Expand Down
5 changes: 2 additions & 3 deletions nvidia_deeplearningexamples_ssd.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ layout: hub_detail
background-class: hub-background
body-class: hub
title: SSD
summary: Single Shot MultiBox Detector model for object detection
summary: 객체 탐지를 위한 Single Shot MultiBox Detector 모델
category: researchers
image: nvidia_logo.png
author: NVIDIA
Expand All @@ -30,8 +30,7 @@ SSD300 모델은 "단일 심층 신경망을 사용하여 이미지에서 물체
* conv4_x의 모든 strides는 1x1로 설정됩니다.

백본 뒤에는 5개의 합성곱 레이어가 추가됩니다. 또한 합성곱 레이어 외에도 6개의 detection heads를 추가했습니다.
The backbone is followed by 5 additional convolutional layers.
In addition to the convolutional layers, we attached 6 detection heads:

* 첫 번째 detection head는 마지막 conv4_x 레이어에 연결됩니다.
* 나머지 5개의 detection head는 추가되는 5개의 합성곱 레이어에 부착됩니다.

Expand Down
2 changes: 1 addition & 1 deletion nvidia_deeplearningexamples_tacotron2.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ layout: hub_detail
background-class: hub-background
body-class: hub
title: Tacotron 2
summary: The Tacotron 2 model for generating mel spectrograms from text
summary: 텍스트에서 멜 스펙트로그램(mel spectrogram)을 생성하는 Tacotron 2 모델
category: researchers
image: nvidia_logo.png
author: NVIDIA
Expand Down
2 changes: 1 addition & 1 deletion nvidia_deeplearningexamples_waveglow.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ layout: hub_detail
background-class: hub-background
body-class: hub
title: WaveGlow
summary: WaveGlow model for generating speech from mel spectrograms (generated by Tacotron2)
summary: 멜 스펙트로그램으로부터 음성을 생성하기 위한 WaveGlow 모델 (Tacotron2에 의해 생성됨)
category: researchers
image: nvidia_logo.png
author: NVIDIA
Expand Down
2 changes: 1 addition & 1 deletion pytorch_vision_alexnet.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ layout: hub_detail
background-class: hub-background
body-class: hub
title: AlexNet
summary: The 2012 ImageNet winner achieved a top-5 error of 15.3%, more than 10.8 percentage points lower than that of the runner up.
summary: 2012년 ImageNet 우승자는 15.3%의 top-5 에러율을 달성하여 준우승자보다 10.8%P 이상 낮았습니다.
category: researchers
image: alexnet2.png
author: Pytorch Team
Expand Down
2 changes: 1 addition & 1 deletion pytorch_vision_deeplabv3_resnet101.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ layout: hub_detail
background-class: hub-background
body-class: hub
title: Deeplabv3
summary: DeepLabV3 models with ResNet-50, ResNet-101 and MobileNet-V3 backbones
summary: ResNet-50, ResNet-101 또는 MobileNet-V3 백본이 포함된 DeepLab V3 모델
category: researchers
image: deeplab2.png
author: Pytorch Team
Expand Down
2 changes: 1 addition & 1 deletion pytorch_vision_densenet.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ layout: hub_detail
background-class: hub-background
body-class: hub
title: Densenet
summary: Dense Convolutional Network (DenseNet), connects each layer to every other layer in a feed-forward fashion.
summary: DenseNet(Dense Convolutional Network)은 피드포워드 방식으로 각 계층을 다른 모든 계층에 연결합니다.
category: researchers
image: densenet1.png
author: Pytorch Team
Expand Down
2 changes: 1 addition & 1 deletion pytorch_vision_fcn_resnet101.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ layout: hub_detail
background-class: hub-background
body-class: hub
title: FCN
summary: Fully-Convolutional Network model with ResNet-50 and ResNet-101 backbones
summary: ResNet-50 ResNet-101 백본을 사용하는 완전 컨볼루션 네트워크 모델
category: researchers
image: fcn2.png
author: Pytorch Team
Expand Down
2 changes: 1 addition & 1 deletion pytorch_vision_ghostnet.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ layout: hub_detail
background-class: hub-background
body-class: hub
title: GhostNet
summary: Efficient networks by generating more features from cheap operations
summary: 적은 연산에서 더 많은 특징을 생성하여 효율적인 네트워크
category: researchers
image: ghostnet.png
author: Huawei Noah's Ark Lab
Expand Down
2 changes: 1 addition & 1 deletion pytorch_vision_googlenet.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ layout: hub_detail
background-class: hub-background
body-class: hub
title: GoogLeNet
summary: GoogLeNet was based on a deep convolutional neural network architecture codenamed "Inception" which won ImageNet 2014.
summary: GoogLeNet은 "Inception"이라는 심층 컨볼루션 신경망 아키텍처를 기반으로 하여 ImageNet 2014에서 수상했습니다.
category: researchers
image: googlenet1.png
author: Pytorch Team
Expand Down
2 changes: 1 addition & 1 deletion pytorch_vision_ibnnet.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ layout: hub_detail
background-class: hub-background
body-class: hub
title: IBN-Net
summary: Networks with domain/appearance invariance
summary: 도메인/외관 불변성을 갖는 네트워크
category: researchers
image: ibnnet.png
author: Xingang Pan
Expand Down
2 changes: 1 addition & 1 deletion pytorch_vision_inception_v3.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ layout: hub_detail
background-class: hub-background
body-class: hub
title: Inception_v3
summary: Also called GoogleNetv3, a famous ConvNet trained on Imagenet from 2015
summary: GoogleNetv3이라고도 불리는, 2015년 ImageNet으로 훈련된 유명한 ConvNet
category: researchers
image: inception_v3.png
author: Pytorch Team
Expand Down
2 changes: 1 addition & 1 deletion pytorch_vision_meal_v2.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ layout: hub_detail
background-class: hub-background
body-class: hub
title: MEAL_V2
summary: Boosting Tiny and Efficient Models using Knowledge Distillation.
summary: Knowledge Distillation을 사용한 작고 효율적인 모델
category: researchers
image: MEALV2.png
author: Carnegie Mellon University
Expand Down
2 changes: 1 addition & 1 deletion pytorch_vision_proxylessnas.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ layout: hub_detail
background-class: hub-background
body-class: hub
title: ProxylessNAS
summary: Proxylessly specialize CNN architectures for different hardware platforms.
summary: 다양한 하드웨어 플랫폼을 위해 프록시 없이 전문화시킨 CNN 아키텍처
category: researchers
image: proxylessnas.png
author: MIT Han Lab
Expand Down
2 changes: 1 addition & 1 deletion pytorch_vision_resnest.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ layout: hub_detail
background-class: hub-background
body-class: hub
title: ResNeSt
summary: A new ResNet variant.
summary: 새로운 ResNet 변형 모델
category: researchers
image: resnest.jpg
author: Hang Zhang
Expand Down
4 changes: 2 additions & 2 deletions pytorch_vision_resnext.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ layout: hub_detail
background-class: hub-background
body-class: hub
title: ResNext
summary: Next generation ResNets, more efficient and accurate
summary: 보다 효율적이고 정확한 차세대 ResNets
category: researchers
image: resnext.png
author: Pytorch Team
Expand Down Expand Up @@ -85,7 +85,7 @@ for i in range(top5_prob.size(0)):

### 모델 설명

Resnext 모델은 논문 [Aggregated Residual Transformations for Deep Neural Networks]에서 제안되었습니다. (https://arxiv.org/abs/1611.05431).
Resnext 모델은 논문 ["Aggregated Residual Transformations for Deep Neural Networks"](https://arxiv.org/abs/1611.05431) 에서 제안되었습니다.
여기서는 50개의 계층과 101개의 계층을 가지는 2개의 resnet 모델을 제공하고 있습니다.
resnet50과 resnext50의 아키텍처 차이는 논문의 Table 1을 참고하십시오.
ImageNet 데이터셋에 대한 사전훈련된 모델의 에러(성능)은 아래 표와 같습니다.
Expand Down
2 changes: 1 addition & 1 deletion pytorch_vision_shufflenet_v2.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ layout: hub_detail
background-class: hub-background
body-class: hub
title: ShuffleNet v2
summary: An efficient ConvNet optimized for speed and memory, pre-trained on Imagenet
summary: ImageNet에서 사전 훈련된 속도와 메모리에 최적화된 효율적인 ConvNet
category: researchers
image: shufflenet_v2_1.png
author: Pytorch Team
Expand Down
2 changes: 1 addition & 1 deletion pytorch_vision_squeezenet.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ layout: hub_detail
background-class: hub-background
body-class: hub
title: SqueezeNet
summary: Alexnet-level accuracy with 50x fewer parameters.
summary: 50배 적은 파라미터로 Alexnet 수준의 정확도 제공
category: researchers
image: squeezenet.png
author: Pytorch Team
Expand Down
6 changes: 3 additions & 3 deletions pytorch_vision_vgg.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ layout: hub_detail
background-class: hub-background
body-class: hub
title: vgg-nets
summary: Award winning ConvNets from 2014 Imagenet ILSVRC challenge
summary: 2014 Imagenet ILSVRC 챌린지에서 ConvNets부분 수상
category: researchers
image: vgg.png
author: Pytorch Team
Expand Down Expand Up @@ -90,12 +90,12 @@ for i in range(top5_prob.size(0)):

### 모델 설명

각 구성 및 bachnorm 버전에 대해서 [Very Deep Convolutional Networks for Large-Scale Image Recognition](https://arxiv.org/abs/1409.1556)에서 제안한 모델에 대한 구현이 있습니다.
각 구성 및 batchnorm 버전에 대해서 [Very Deep Convolutional Networks for Large-Scale Image Recognition](https://arxiv.org/abs/1409.1556)에서 제안한 모델에 대한 구현이 있습니다.

예를 들어, 논문에 제시된 구성 `A`는 `vgg11`, `B`는 `vgg13`, `D`는 `vgg16`, `E`는 `vgg19`입니다.
batchnorm 버전은 `_bn`이 접미사로 붙어있습니다.

사전 훈련된 모델이 있는 imagenet 데이터 세트의 1-crop 오류율은 아래에 나열되어 있습니다.
사전 훈련된 모델이 있는 ImageNet 데이터 세트의 Top-1 오류율은 아래에 나열되어 있습니다.

| Model structure | Top-1 error | Top-5 error |
| --------------- | ----------- | ----------- |
Expand Down
Loading