Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat
Sign Up
Cited by 262

The DOI matches its Crossref record; 11 of the 11 references it cites were found in a registry or an index.

Listen to Look: Action Recognition by Previewing Audio

Open access
Gao, Ruohan, Oh, Tae-Hyun, Grauman, Kristen, Torresani, Lorenzo·2020·2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
DOI 10.1109/cvpr42600.2020.01047
Open this paper in CoChat — free account

Abstract

In the face of the video data deluge, today's expensive clip-level classifiers are increasingly impractical. We propose a framework for efficient action recognition in untrimmed video that uses audio as a preview mechanism to eliminate both short-term and long-term visual redundancies. First, we devise an ImgAud2Vid framework that hallucinates clip-level features by distilling from lighter modalities---a single frame and its accompanying audio---reducing short-term temporal redundancy for efficient clip-level recognition. Second, building on ImgAud2Vid, we further propose ImgAud-Skimming, an attention-based long short-term memory network that iteratively selects useful moments in untrimmed videos, reducing long-term temporal redundancy for efficient video-level recognition. Extensive experiments on four action recognition datasets demonstrate that our method achieves the state-of-the-art in terms of both recognition accuracy and speed.

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1109/cvpr42600.2020.01047

Cites

11 of 11 references verified0 citation errors · 0 not checkable

Statistics

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 11 references

Research works printed but not linked: 78

  1. 10.1109/CVPR.2019.00037 · DOI · Paper page
  2. 10.1007/978-3-540-88688-4_48 · DOI · Paper page
  3. wang, 2019, What makes training multi-modal networks hard?, ArXiv Preprint
  4. 10.1145/2964284.2964328 · DOI
  5. 10.1109/CVPR.2019.00137 · DOI
  6. 10.1109/CVPR.2018.00631 · DOI · Paper page
  7. 10.1109/ICCV.2019.00718 · DOI · Paper page
  8. 10.1109/ICCV.2019.00632 · DOI
  9. lee, 2012, Discovering important people and objects for egocentric video summarization, CVPR
  10. xie, 2018, Rethinking spatiotemporal feature learning: Speed-accuracy trade-offs in video classification, ECCV
  11. 10.1109/ICCV.2017.617 · DOI · Paper page
  12. kay, 2017, The kinetics human action video dataset, ArXiv Preprint
  13. 10.1109/CVPR.2014.100 · DOI
  14. 10.1109/CVPR.2013.332 · DOI
  15. 10.1109/ICCV.2003.1238378 · DOI · Paper page
  16. 10.1109/ICCV.2019.00633 · DOI
  17. korbar, 2018, Co-training of audio and video representations from self-supervised temporal synchronization, NeurIPS
  18. 10.1109/ICCV.2019.00559 · DOI · Paper page
  19. tian, 2018, Audio-visual event localization in unconstrained videos, ECCV
  20. 10.1109/CVPR.2018.00675 · DOI
  21. varol, 2017, Long-term temporal convolutions for action recognition, TPAMI
  22. vaswani, 2017, Attention is all you need, NeurIPS
  23. vinyals, 2015, Pointer networks, NeurIPS
  24. hinton, 2015, Distilling the knowledge in a neural network, ArXiv Preprint
  25. 10.1109/CVPR.2013.345 · DOI
  26. 10.1109/CVPR.2017.678 · DOI · Paper page
  27. wang, 2016, Temporal segment networks: Towards good practices for deep action recognition, ECCV
  28. 10.21437/Interspeech.2018-1400 · DOI
  29. 2017, Cisco visual networking index: Forecast and trends, White Paper
  30. gao, 2018, Learning to separate object sounds by watching unlabeled video, ECCV
  31. 10.1109/ICCV.2019.00398 · DOI
  32. 10.1109/CVPR.2019.00041 · DOI
  33. 10.1109/CVPR.2017.337 · DOI · Paper page
  34. 10.1207/s15326969eco0501_1 · DOI · Paper page
  35. gong, 2014, Diverse sequential subset selection for supervised video summarization, NeurIPS
  36. girdhar, 2019, Dis-tinit: Learning video representations without a single labeled video, ICCV
  37. 10.1109/CVPR.2012.6247807 · DOI
  38. 10.1109/CVPR.2018.00474 · DOI · Paper page
  39. 10.1109/ICCV.2019.00756 · DOI · Paper page
  40. sukhbaatar, 2015, End-to-end memory networks, NeurIPS
  41. 10.1007/978-3-319-46478-7_48 · DOI
  42. soomro, 2012, Ucf101: A dataset of 101 human actions classes from videos in the wild, ArXiv Preprint
  43. simonyan, 2014, Two-stream convolutional networks for action recognition in videos, NeurIPS
  44. 10.1109/CVPR.2019.00136 · DOI
  45. 10.1109/CVPR.2017.155 · DOI · Paper page
  46. 10.1109/TPAMI.2019.2952095 · DOI
  47. chen, 2018, Multi-fiber networks for video recognition, ECCV
  48. lin, 2018, Bsn: Boundary sensitive network for temporal action proposal generation, ECCV
  49. 10.1109/CVPR.2015.7298878 · DOI · Paper page
  50. 10.1109/CVPR.2019.00190 · DOI
  51. 10.1145/3197517.3201357 · DOI · Paper page
  52. 10.24963/ijcai.2018/98 · DOI
  53. 10.1007/978-3-030-01237-3_24 · DOI
  54. 10.1109/CVPR.2015.7299101 · DOI · Paper page
  55. 10.1109/CVPR.2016.213 · DOI · Paper page
  56. 10.1109/ICCV.2017.317 · DOI · Paper page
  57. 10.1109/CVPR.2015.7299176 · DOI · Paper page
  58. zhao, 2018, The sound of pixels, ECCV
  59. 10.1109/ICCV.2019.00715 · DOI
  60. zolfaghari, 2018, Eco: Efficient convolutional network for online video understanding, ECCV
  61. alwassel, 2018, Action search: Spotting actions in videos and its application to temporal action localization, ECCV
  62. 10.1145/3240508.3240578 · DOI · Paper page
  63. arandjelovi?, 2018, Objects that sound, ECCV
  64. 10.1109/ICCV.2017.73 · DOI · Paper page
  65. zhou, 2018, Temporal relational reasoning in videos, ECCV
  66. 10.1109/CVPR.2017.675 · DOI · Paper page
  67. 10.1109/ICCV.2019.00037 · DOI
  68. aytar, 2016, Soundnet: Learning sound representations from unlabeled video, NeurIPS
  69. 10.1109/ICCV.2017.590 · DOI · Paper page
  70. 10.1109/CVPR.2018.00374 · DOI · Paper page
  71. 10.1007/978-3-030-01228-1_9 · DOI
  72. 10.1109/CVPR.2016.264 · DOI · Paper page
  73. 10.1109/CVPR.2014.85 · DOI
  74. owens, 2016, Ambient sound provides supervision for visual learning, ECCV
  75. 10.1109/CVPR.2017.318 · DOI · Paper page
  76. 10.1109/CVPR.2018.00817 · DOI
  77. morgado, 2018, Self-supervised generation of spatial audio for 360° video, NeurIPS
  78. mnih, 2014, Recurrent models of visual attention, NeurIPS

Cited by

See all 262

Metadata

Authors
Gao, R., Oh, T.-H., Grauman, K., & Torresani, L.
Venue
2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
Publisher
IEEE
Published
2020-06
Pages
10454-10464
ISSN
1063-6919 (ISSN-L)
DOI
10.1109/cvpr42600.2020.01047
OpenAlex
W3034658206
Cited by
262
References
11
FWCI
13.79
Type
conference paper
Language
en
Open access
green
Licence

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Listen to Look: Action Recognition by Previewing Audio. https://app.cochat.ai/papers/10.1109/cvpr42600.2020.01047
The paper
Gao, R., Oh, T.-H., Grauman, K., & Torresani, L. (2020). Listen to Look: Action Recognition by Previewing Audio. 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). https://doi.org/10.1109/cvpr42600.2020.01047

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.