Aller au contenu

ModelPrésentéExterne

CAVIS

Seunghun Lee, Jiwan Seo, Kiljoon Han, Minwoo Choi, Sunghoon Im (DGIST)

Sourcehuggingface.co/DGIST-CVLAB-Video/CAVIS/tree/4973efd2fdda29adc238db927737baac1f69e9bc

TypeOtherVersionRevision 4973efdCClasse

Source du modèle

Type: Other
Taille du fichier: 169 MB–2,42 GB
Version: vRevision 4973efd
Mis à jour: 15 oct. 2026
Téléchargements: 0
Source: Hugging Face

Description de la source

CAVIS: Context-Aware Video Instance Segmentation (ICCV 2025)

Seunghun Lee, Jiwan Seo, Kiljoon Han, Minwoo Choi, Sunghoon Im (DGIST)

Link to Github
Link to Paper

Model Details

Model Architecture

Model Architecture

Model Sources [optional]

Repository:
**Paper:
**Demo:
**Github:

How to Get Started with the Model

Evaluation

Results

Sources

1 source

Hugging FacePrincipalVersion 4973efd2fdda29adc238db927737baac1f69e9bc

huggingface.co/DGIST-CVLAB-Video/CAVIS/tree/4973efd2fdda29adc238db927737baac1f69e9bc

Vérifié 28 août

Artefacts du modèle

27 artefacts

CAVIS_offline_ovis_ViTL_57.1.pth

pth · 1,39 GB · SHA-256 0dc9977d772c…cc91 · Hugging Face

CAVIS_offline_VIPSeg_R50_45.3.pth

pth · 262 MB · SHA-256 243cf1115fbb…88a1 · Hugging Face

CAVIS_offline_VIPSeg_ViTL_58.5.pth

pth · 1,39 GB · SHA-256 3d2e511cdea8…4535 · Hugging Face

CAVIS_offline_ytvis19_ViTL_69.4.pth

pth · 1,39 GB · SHA-256 1bb6e9047232…6a60 · Hugging Face

CAVIS_offline_ytvis21_ViTL_65.3.pth

pth · 1,39 GB · SHA-256 f7d5808a3ef8…c568 · Hugging Face

CAVIS_online_ovis_R50_37.6.pth

pth · 282 MB · SHA-256 16038bb821a2…6113 · Hugging Face

CAVIS_online_ovis_SwinL_48.6.pth

pth · 865 MB · SHA-256 208280a68012…ae76 · Hugging Face

CAVIS_online_ovis_ViTL_53.2.pth

pth · 1,33 GB · SHA-256 490a8789b6ae…7252 · Hugging Face

CAVIS_online_VIPSeg_R50_42.4.pth

pth · 282 MB · SHA-256 6b20508d0615…8d21 · Hugging Face

CAVIS_online_VIPSeg_ViTL_56.9.pth

pth · 1,33 GB · SHA-256 924acd5e4192…1c32 · Hugging Face

CAVIS_online_ytvis19_R50_55.7.pth

pth · 207 MB · SHA-256 3849e3f16884…3ff7 · Hugging Face

CAVIS_online_ytvis19_SwinL_66.0.pth

pth · 865 MB · SHA-256 2294e5dcbe46…93c4 · Hugging Face

CAVIS_online_ytvis19_ViTL_68.9.pth

pth · 1,41 GB · SHA-256 72efa63ceea4…a438 · Hugging Face

CAVIS_online_ytvis21_R50_50.5.pth

pth · 207 MB · SHA-256 233dd1ff819d…7427 · Hugging Face

CAVIS_online_ytvis21_SwinL_61.1.pth

pth · 865 MB · SHA-256 5ba18d28f6a9…11de · Hugging Face

CAVIS_online_ytvis21_ViTL_64.6.pth

pth · 1,33 GB · SHA-256 eb6e9fcd92c7…97bb · Hugging Face

CAVIS_segmenter_ovis_R50_30.0.pth

pth · 507 MB · SHA-256 bb16c1f19cda…0c51 · Hugging Face

CAVIS_segmenter_ovis_SwinL_41.2.pth

pth · 2,42 GB · SHA-256 076667877b84…5317 · Hugging Face

CAVIS_segmenter_ovis_ViTL_45.5.pth

pth · 1,30 GB · SHA-256 f8696f57a60a…fb4a · Hugging Face

CAVIS_segmenter_VIPSeg_R50_40.7.pth

pth · 507 MB · SHA-256 bcccde77cb55…a594 · Hugging Face

CAVIS_segmenter_VIPSeg_ViTL_56.1.pth

pth · 1,30 GB · SHA-256 d0919a5da047…11a1 · Hugging Face

CAVIS_segmenter_ytvis19_R50_51.9.pth

pth · 507 MB · SHA-256 9bd7fead454c…a7be · Hugging Face

CAVIS_segmenter_ytvis19_SwinL_63.2.pth

pth · 827 MB · SHA-256 08adf833905c…cd70 · Hugging Face

CAVIS_segmenter_ytvis19_ViTL_66.7.pth

pth · 1,63 GB · SHA-256 5f1ca6ab7dcf…c4f1 · Hugging Face

CAVIS_segmenter_ytvis21_R50_47.4.pth

pth · 169 MB · SHA-256 f05e1b9abb18…6257 · Hugging Face

CAVIS_segmenter_ytvis21_SwinL_57.8.pth

pth · 827 MB · SHA-256 d83d8c02e257…f395 · Hugging Face

CAVIS_segmenter_ytvis21_ViTL_61.7.pth

pth · 1,63 GB · SHA-256 27ce560a9b74…2921 · Hugging Face

Extraits de sources

2 extraits

Readme28 aoûthuggingface.co/DGIST-CVLAB-Video/CAVIS

--- license: mit language: - en pipeline_tag: video-classification --- # CAVIS: Context-Aware Video Instance Segmentation (ICCV 2025) Seunghun Lee, Jiwan Seo, Kiljoon Han, Minwoo Choi, Sunghoon Im (DGIST) Link to [Github](https://github.com/Seung-Hun-Lee/CAVIS) \ Link to [Paper](https://arxiv.org/abs/2407.03010) ## Model Details ![Model Architecture](images/CAVIS.PNG) ![Model Architecture](images/Method.png) ### Model Sources [optional]  - **Repository:** - **Paper: - **Demo: - **Github: ## How to Get Started with the Model ## Evaluation ### Results

Note28 aoûthuggingface.co/DGIST-CVLAB-Video/CAVIS

Source context: 0 downloads · 1 likes · Pipeline video-classification · Repo DGIST-CVLAB-Video/CAVIS