Downloads · 30 days
0
xieyuy/traffic-sign-adverse-weather
traffic-sign-adverse-weather is a image classification model from xieyuy. Use it when you need a label for an image. It is set up for timm. The card lists the license as mit.
Official model checkpoints for the solution in the Traffic Sign Recognition under Adverse Weather Competition.
Downloads · 30 days
0
Access
Public
Updated Sep 18, 2026
Repo size
913 MB
Likes
1
Public
Click a slice to open those files.
.pth913 MB · 100%
From the Hugging Face model README
Official model checkpoints for the solution in the Traffic Sign Recognition under Adverse Weather Competition.
| Model | Checkpoint File | Parameters | Input Resolution | Val Macro-F1 |
|---|---|---|---|---|
| ConvNeXt V2-Base | convnextv2_base_best.pth | 89M | 384x384 | 0.938 |
| Swin Transformer V2-Base | swin_v2_b_best.pth | 88M | 256x256 | 0.916 |
| EfficientNetV2-M | efficientnet_v2_m_best.pth | 54M | 384x384 | 0.917 |
See classes.txt for the 25 traffic sign classes.
from huggingface_hub import hf_hub_download
# Download ConvNeXt V2-Base checkpoint
checkpoint_path = hf_hub_download(
repo_id="xieyuy/traffic-sign-adverse-weather",
filename="convnextv2_base_best.pth",
local_dir="./results"
)
For inference scripts, training code, and in-depth engineering retrospective, visit the GitHub Repository.