Zobrazeno 1 - 9
of 9
pro vyhledávání: '"Miech, Antoine"'
Autor:
Yang, Antoine, Nagrani, Arsha, Seo, Paul Hongsuck, Miech, Antoine, Pont-Tuset, Jordi, Laptev, Ivan, Sivic, Josef, Schmid, Cordelia
Publikováno v:
CVPR 2023-IEEE/CVF Conference on Computer Vision and Pattern Recognition
CVPR 2023-IEEE/CVF Conference on Computer Vision and Pattern Recognition, Jun 2023, Vancouver, Canada
CVPR 2023-IEEE/CVF Conference on Computer Vision and Pattern Recognition, Jun 2023, Vancouver, Canada
In this work, we introduce Vid2Seq, a multi-modal single-stage dense event captioning model pretrained on narrated videos which are readily-available at scale. The Vid2Seq architecture augments a language model with special time tokens, allowing it t
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::e67c8cfe1931fcccfa44cf37962312ff
Publikováno v:
NeurIPS 2022-36th Conference on Neural Information Processing Systems
NeurIPS 2022-36th Conference on Neural Information Processing Systems, Nov 2022, New Orleans, United States
NeurIPS 2022-36th Conference on Neural Information Processing Systems, Nov 2022, New Orleans, United States
Video question answering (VideoQA) is a complex task that requires diverse multi-modal data for training. Manual annotation of question and answers for videos, however, is tedious and prohibits scalability. To tackle this problem, recent methods cons
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::756e8bcf78e55b07afbe61778dc057a1
http://arxiv.org/abs/2206.08155
http://arxiv.org/abs/2206.08155
Publikováno v:
CVPR 2022-IEEE/CVF Conference on Computer Vision and Pattern Recognition
CVPR 2022-IEEE/CVF Conference on Computer Vision and Pattern Recognition, Jun 2022, New Orleans, United States
CVPR 2022-IEEE/CVF Conference on Computer Vision and Pattern Recognition, Jun 2022, New Orleans, United States
We consider the problem of localizing a spatio-temporal tube in a video corresponding to a given text query. This is a challenging task that requires the joint and efficient modeling of temporal, spatial and multi-modal interactions. To address this
Publikováno v:
ICCV 2021-IEEE International Conference on Computer Vision
ICCV 2021-IEEE International Conference on Computer Vision, Oct 2021, Montréal, Canada
ICCV 2021-IEEE International Conference on Computer Vision, Oct 2021, Virtual, France
ICCV 2021-IEEE International Conference on Computer Vision, Oct 2021, Montréal, Canada
ICCV 2021-IEEE International Conference on Computer Vision, Oct 2021, Virtual, France
Recent methods for visual question answering rely on large-scale annotated datasets. Manual annotation of questions and answers for videos, however, is tedious, expensive and prevents scalability. In this work, we propose to avoid manual annotation a
Autor:
Miech, Antoine
Publikováno v:
Computer Vision and Pattern Recognition [cs.CV]. PSL Research University, 2020. English
The goal of this thesis is to build and train machine learning models capable of understanding the content of videos. Current video understanding approaches mainly rely on large-scale manually annotated video datasets for training. However, collectin
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=dedup_wf_001::245a8f3d7568540c31ffddbdc000bcb5
https://hal.inria.fr/tel-03084216
https://hal.inria.fr/tel-03084216
Autor:
Miech, Antoine
Publikováno v:
Computer Vision and Pattern Recognition [cs.CV]. PSL Research University, 2020. English
Computer Vision and Pattern Recognition [cs.CV]. Université Paris sciences et lettres, 2020. English. ⟨NNT : 2020UPSLE059⟩
Computer Vision and Pattern Recognition [cs.CV]. PSL Research University, 2020. English. ⟨NNT : ⟩
Computer Vision and Pattern Recognition [cs.CV]. Université Paris sciences et lettres, 2020. English. ⟨NNT : 2020UPSLE059⟩
Computer Vision and Pattern Recognition [cs.CV]. PSL Research University, 2020. English. ⟨NNT : ⟩
The goal of this thesis is to build and train machine learning models capable of understanding the content of videos. Current video understanding approaches mainly rely on large-scale manually annotated video datasets for training. However, collectin
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=dedup_wf_001::78f4bdb76ee4297241e7bdbfa5ef4426
https://hal.inria.fr/tel-03084216/file/main.pdf
https://hal.inria.fr/tel-03084216/file/main.pdf
Publikováno v:
IEEE Transactions on Pattern Analysis and Machine Intelligence
IEEE Transactions on Pattern Analysis and Machine Intelligence, 2022, ⟨10.1109/tpami.2022.3173208⟩
IEEE Transactions on Pattern Analysis and Machine Intelligence, 2022, ⟨10.1109/tpami.2022.3173208⟩
Recent methods for visual question answering rely on large-scale annotated datasets. Manual annotation of questions and answers for videos, however, is tedious, expensive and prevents scalability. In this work, we propose to avoid manual annotation a
Current methods for video analysis often extract frame-level features using pre-trained convolutional neural networks (CNNs). Such features are then aggregated over time e.g., by simple temporal averaging or more sophisticated recurrent neural networ
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::4386952ac64cd6917aa7ab0cec0861a1
https://hal.inria.fr/hal-01547378/document
https://hal.inria.fr/hal-01547378/document
Autor:
Jean-Baptiste Alayrac, Josef Sivic, Antoine Miech, Ivan Laptev, Dimitri Zhukov, Makarand Tapaswi
Publikováno v:
ICCV 2019-International Conference on Computer Vision
ICCV 2019-International Conference on Computer Vision, Oct 2019, Séoul, South Korea
ICCV
ICCV 2019-International Conference on Computer Vision, Oct 2019, Séoul, South Korea
ICCV
Learning text-video embeddings usually requires a dataset of video clips with manually provided captions. However, such datasets are expensive and time consuming to create and therefore difficult to obtain on a large scale. In this work, we propose i
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::720f44a13124db3df7f8d3e8279c7bc2
https://hal.science/hal-02433497
https://hal.science/hal-02433497