Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
一般物体検出とLSTMを用いた画像に基づく屋内位置推定 - IPSJ UBI82
Search
Aokiti
May 12, 2024
490
0
Share
一般物体検出とLSTMを用いた画像に基づく屋内位置推定 - IPSJ UBI82
http://id.nii.ac.jp/1001/00233750/
Aokiti
May 12, 2024
More Decks by Aokiti
See All by Aokiti
d-hacks 今期運営 2025f
sakusaku3939
0
410
[d-hacks Docker講座] Dockerで動かすローカルLLM入門
sakusaku3939
0
64
[論文輪読会] A survey of model compression strategies for object detection
sakusaku3939
0
30
[論文輪読会] ViT-1.58b
sakusaku3939
0
200
d-hacks PyTorchモデル実装会 2024f
sakusaku3939
0
53
[論文輪読会] Binarized Neural Networks
sakusaku3939
0
62
MoodTune 東京AI祭ハッカソン決勝
sakusaku3939
0
610
d-hacks PyTorch実装会 2023f
sakusaku3939
0
31
[DL勉強会] 第5章 ディープラーニングを活用したアプリケーション 後半
sakusaku3939
0
18
Featured
See All Featured
Deep Space Network (abreviated)
tonyrice
0
150
State of Search Keynote: SEO is Dead Long Live SEO
ryanjones
0
190
Are puppies a ranking factor?
jonoalderson
1
3.4k
The Straight Up "How To Draw Better" Workshop
denniskardys
239
140k
HU Berlin: Industrial-Strength Natural Language Processing with spaCy and Prodigy
inesmontani
PRO
0
380
SEO in 2025: How to Prepare for the Future of Search
ipullrank
3
3.4k
HTML-Aware ERB: The Path to Reactive Rendering @ RubyCon 2026, Rimini, Italy
marcoroth
1
54
Building an army of robots
kneath
306
46k
Save Time (by Creating Custom Rails Generators)
garrettdimon
PRO
32
3k
Agile that works and the tools we love
rasmusluckow
331
21k
I Don’t Have Time: Getting Over the Fear to Launch Your Podcast
jcasabona
34
2.7k
It's Worth the Effort
3n
188
29k
Transcript
1
2
3
4
5
6
[7] F. Walch, C. Hazirbas, L. Leal-Taixé, T. Sattler, S.
Hilsenbeck, D. Cremers 7 https://doi.org/10.1109/ICCV.2017.75
8 [8] S Nilwong, D Hossain, S Kaneko, G Capi
https://doi.org/10.3390/machines7020025
9
10 CNN(GoogLeNetモデル) LSTM(次元を削減) CNN CNN CNN LSTM(1つの特徴量に変換)
11 CNN CNN CNN LSTM(1つの特徴量に変換)
12
13
14
15
16 𝑀𝐴𝐸 = 1 𝑛 𝑖=1 𝑛 |ෝ 𝑥𝑖
− 𝑥𝑖 + | ෝ 𝑦𝑖 − 𝑦𝑖 |)
17 CNN (4層) CNN(4層) CNN(4層) LSTM(1つの特徴量に変換) CNN(4層) CNN(GoogLeNet) CNN(GoogLeNet) LSTM(次元を削減)
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34