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
Taking photos is free
Search
Xyggy
February 21, 2019
Technology
0
440
Taking photos is free
Xyggy
February 21, 2019
Tweet
Share
More Decks by Xyggy
See All by Xyggy
Add and remove images from query
xyggy
0
920
More like this or these
xyggy
0
900
More of these, less of the others
xyggy
0
870
Knowns and unknowns
xyggy
0
390
Realtime AI is Not a Myth
xyggy
0
1.6k
Litmus test
xyggy
0
830
3 minute Thingy
xyggy
0
110
pop-your-filter-bubble.pdf
xyggy
0
210
Other Decks in Technology
See All in Technology
過去のインプットとアウトプットを振り返る
diggymo
0
110
ドメインと向き合う - 旅行予約編
hidenorigoto
4
550
テストコードの品質を客観的な数値で担保しよう〜Mutation Testのすすめ〜
ysknsid25
10
3k
Perlで始めるeBPF: 自作Loaderの作り方 / Getting started with eBPF in Perl_How to create your own Loader
takehaya
1
770
AI時代のアジャイル開発(XP祭り2024版) / Agile Development in the AI Era in XPJUG
takaking22
13
3.6k
OPENLOGI Company Profile for engineer
hr01
1
12k
スクラム導入の舞台裏:QAエンジニアがスクラムマスターになるまで
bubo1201
0
170
小さな勉強会の始め方、広げ方、あるいは友達の作り方 / How to Start, Grow, and Build Connections with Small Study Groups
ar_tama
5
2.4k
All your memory are belong to… whom?
ennael
PRO
0
630
How CERN serves 1EB of data via FUSE
ennael
PRO
0
16k
これはPerl? それともRuby? クイズ〜〜〜〜〜!!!- Perl or Ruby Quiz
moznion
2
1.6k
Pythonを活用したLLMによる構造的データ生成の手法と実践
brainpadpr
3
290
Featured
See All Featured
Visualization
eitanlees
143
15k
GitHub's CSS Performance
jonrohan
1030
450k
The Cult of Friendly URLs
andyhume
77
6k
10 Git Anti Patterns You Should be Aware of
lemiorhan
653
59k
Fireside Chat
paigeccino
32
3k
Creating an realtime collaboration tool: Agile Flush - .NET Oxford
marcduiker
24
1.7k
Evolution of real-time – Irina Nazarova, EuRuKo, 2024
irinanazarova
3
220
Documentation Writing (for coders)
carmenintech
65
4.3k
Java REST API Framework Comparison - PWX 2021
mraible
PRO
28
7.5k
Save Time (by Creating Custom Rails Generators)
garrettdimon
PRO
25
660
Done Done
chrislema
181
16k
Stop Working from a Prison Cell
hatefulcrawdad
267
20k
Transcript
1 Taking photos is free Farmer wants to know disease
type affecting tomato crop. Photos are from the same plant. Dataset: 80K plant leaf images of different crops https://plantvillage.org/ No image pre-processing performed. No textual data used. Feature vectors generated automatically with deep learning. #1 #2 #3 www.xyggy.com © 2019
2 Query with photo #1 Top 5 results: 1 potato_early_blight
2 tomato_early_blight 3 potato_early_blight 4 potato_early_blight 5 tomato_early_blight 2nd result identifies disease www.xyggy.com © 2019
3 Query with photo #2 Top 5 results: 1 tomato_early_blight
2 celery_early_blight 3 tomato_early_blight 4 potato_early_blight 5 tomato_early_blight 1st result identifies disease www.xyggy.com © 2019
4 Query with photo #3 Top 5 results: 1 soybean_leaf_blight
2 celery_early_blight 3 potato_early_blight 4 tomato_early_blight 5 soybean_leaf_blight 4th result identifies disease www.xyggy.com © 2019
5 Real world data is noisy 3 queries with different
photos from the same tomato plant give different predictions. Real world data is noisy. Let’s try with Thingy using multiple photos per query. www.xyggy.com © 2019
6 Query with photo #1 and #2 Top 5 results:
1 tomato_early_blight 2 tomato_early_blight 3 celery_early_blight 4 potato_early_blight 5 potato_early_blight top 2 positions identify disease www.xyggy.com © 2019
7 Add photo #3 to query top 3 positions identify
disease Top 5 results: 1 tomato_early_blight 2 tomato_early_blight 3 tomato_early_blight 4 potato_early_blight 5 potato_early_blight www.xyggy.com © 2019
8 Taking photos is free Taking photos is free. Ditto
audio, video, sensor and other item data. Multiple items per query gives Thingy additional information to work with to improve relevance. Perfect for a noisy world. www.xyggy.com © 2019
9 Thingy www.xyggy.com © 2019