Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.
2017/7/12
YouTube-8M:
A Large-Scale Video Classification Benchmark
[ Google Research 2016/9/27 arXiv: 1609.08675v1] TFUG N...
✤
✤
✤
✤
✤
✤
✤
✤ Kaggle
2
YouTube-8M
✤
ImageNet…
3
YouTube-8M
✤
4
2TB
1GPU 1
1) 1 1
2) Inception
3) PCA
4) TensorFlow
→ URL
✤ YouTube
5
✤ Knowledge Graph entity
6
✤
✤ 3 1-2.5
7
✤
✤
✤ …
✤ 78.8% 14.5%
✤ → 80%
✤ →
8
DBoF
✤ k N
✤ ReLu
M
✤ → (MxN)
✤
✤ Max pooling
✤
✤
9
✤
✤
✤ φ
✤ PCA
✤ L2
10
✤ mAP:
✤ Hit@k: k
1
✤ PERR(Precision at equal recall rate):
✤ GAP: Kaggle
-
11
✤ 1
✤
✤ 2
DBoF,LSTM
12
→
✤
✤ PERR
15%
✤
13
✤ ActivityNet
✤ Sports-1M
14
Kaggle
✤
✤ 6
✤ Google Cloud …
15
Kaggle 1
✤ https://github.com/antoine77340/LOUPE
✤ Learnable pooling with Context Gating for video classification
✤ [Antoin...
✤ Q.p8
✤ A.
✤
…)
80% 80%
p8
17
Upcoming SlideShare
Loading in …5
×

【NN輪読会】YouTube-8M: A Large-Scale Video Classification Benchmark

1,221 views

Published on

NN論文を肴に酒を飲む会 #3 ②発表資料です

Published in: Data & Analytics
  • I like this service ⇒ www.WritePaper.info ⇐ from Academic Writers. I don't have enough time write it by myself.
       Reply 
    Are you sure you want to  Yes  No
    Your message goes here

【NN輪読会】YouTube-8M: A Large-Scale Video Classification Benchmark

  1. 1. 2017/7/12 YouTube-8M: A Large-Scale Video Classification Benchmark [ Google Research 2016/9/27 arXiv: 1609.08675v1] TFUG NN #3 1
  2. 2. ✤ ✤ ✤ ✤ ✤ ✤ ✤ ✤ Kaggle 2
  3. 3. YouTube-8M ✤ ImageNet… 3
  4. 4. YouTube-8M ✤ 4 2TB 1GPU 1 1) 1 1 2) Inception 3) PCA 4) TensorFlow → URL
  5. 5. ✤ YouTube 5
  6. 6. ✤ Knowledge Graph entity 6
  7. 7. ✤ ✤ 3 1-2.5 7
  8. 8. ✤ ✤ ✤ … ✤ 78.8% 14.5% ✤ → 80% ✤ → 8
  9. 9. DBoF ✤ k N ✤ ReLu M ✤ → (MxN) ✤ ✤ Max pooling ✤ ✤ 9
  10. 10. ✤ ✤ ✤ φ ✤ PCA ✤ L2 10
  11. 11. ✤ mAP: ✤ Hit@k: k 1 ✤ PERR(Precision at equal recall rate): ✤ GAP: Kaggle - 11
  12. 12. ✤ 1 ✤ ✤ 2 DBoF,LSTM 12 →
  13. 13. ✤ ✤ PERR 15% ✤ 13
  14. 14. ✤ ActivityNet ✤ Sports-1M 14
  15. 15. Kaggle ✤ ✤ 6 ✤ Google Cloud … 15
  16. 16. Kaggle 1 ✤ https://github.com/antoine77340/LOUPE ✤ Learnable pooling with Context Gating for video classification ✤ [Antoine Miech arXiv:1706.06905v1 2017/6/21] ✤ 25 ✤ 7 GAP 84.698% Gated NetVLAD (256 clusters), Gated NetFV (128 clusters), Gated Soft-DBoW (4096 clusters), Soft-DBoW (8000 Clusters), Gated NetRVLAD (256 Clusters), GRU (2 layers, hidden size: 1200) LSTM (2 layers, hidden size: 1024) 16
  17. 17. ✤ Q.p8 ✤ A. ✤ …) 80% 80% p8 17

×