Slide 1

Slide 1 text

- - - -

Slide 2

Slide 2 text

Agenda u C u u - u

Slide 3

Slide 3 text

!# (Bernardi et al. 2016) u 1. ! u 2. ! u 3.  ! u  "   1. Patterson, G., Xu, C., Su, H., and Hays, J. The sun attribute databased: Beyond categories for deeper scene understanding. International Journal of Computer Vision 108(1-2):59-81. 2014 2. Devlin, J., Cheng, H., Fang, H., Gupta, S., Deng, L., He, X., Zweig, G., and Mitchell, M. Language models for image captioning: The quirks and what works. arXiv preprint arXiv:1505.01809. 2015 3. Socher, R., Karpathy, A., Le, Q. V., Manning, C. D., and Ng, A. Y. Grounded compositional semantics for finding and describing images with sentences. TACL 2:207–218. 2014 4. Soto, A. J., Kiros, R., Keselj, V., and Milios, E. E. Machine learning meets visualization for extracting insights from text data. AI Matters 2(2):15-17. 2015 5. Karpathy, A., and Fei-Fei, L. Deep visual-semantic alignments for generating image descriptions. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 3128–3137. 2015 6. Donahue, J., Anne Hendricks, L., Guadarrama, S., Rohrbach, M., Venugopalan, S., Saenko, K., and Darrell, T. Long-term recurrent convolutional networks for visual recognition and description. In Proceedings of the IEEE conference on computer vision and pattern recognition, 2625–2634. 2015 7. Schwarz, K., Berg, T. L., and Lensch, H. P. Autoillustrating poems and songs with style. In Asian Conference on Computer Vision, 87–103. 2016

Slide 4

Slide 4 text

From Captions to Image Concepts and Back (Fang et el, 2014) u 2014 MSCOCO rank #1

Slide 5

Slide 5 text

Classic Image caption generator (Vinyals et al, 2014.) Also 2014 MSCOCO rank #1

Slide 6

Slide 6 text

Attention added (Xu et el, 2015.) u Attention weight is calculated from decoder hidden state h and encoder features . u the context vector z is calculated from the attention. u Output and new hidden state is generated from the context vector z and hidden state.

Slide 7

Slide 7 text

Input higher level of image features (Chen et el, 2015.) u Reconstruct visual representation u Stronger connection between images and texts

Slide 8

Slide 8 text

AI  u  u AI     u   

Slide 9

Slide 9 text

  u u  u 

Slide 10

Slide 10 text

1950 Stochastische Texte . 2 2 1 2 . (Liu et al. 2018) P(Wt | wt-1:1 ) =P(w1)*P(w2 |w1 )*P(w3 |w1 w2 )….)

Slide 11

Slide 11 text

/ u u ? u / u / u Perplexity, BLEU score /

Slide 12

Slide 12 text

 u   u   u   https://poem.msxiaobing.com/

Slide 13

Slide 13 text

   AI    

Slide 14

Slide 14 text

-   u " 2!  ($ "%$  ),.$ 3 460$# u ' $1& ($ 5* /$+

Slide 15

Slide 15 text

u      

Slide 16

Slide 16 text

&RK,.1 u !# \QA`L WG4Z?A+NL W>-AE3 BMX @ 0( BW4Z?A+_4 P) A] ^O8UUA;9 !# \QA`L u !# J,.1ISA aR <$2"7DV '=CAT : bA 62"7 [Y95Ad% <$*F/ !# HcA 

Slide 17

Slide 17 text

Case Study: 1 9 ? u : ) 2 u 0 ( 8 ) u 9 5 : 2 9 8

Slide 18

Slide 18 text

7B,1'4-8 u  A=) u A$&@!?< u @!=8 u +,>90C "6,/( # ,;: u +25 ,'3*. %8

Slide 19

Slide 19 text

=0!"6$ 1. 42:9 2. 56) 0-';" 87 0"/ 3. *56 <#%&31, +( .  0   0.975  0.005 - 0.005 0.004 ' 0.003 ( 0.586  0.109  $% 0.053     +    (   !  (% !&% )*+% . #/,%"          

Slide 20

Slide 20 text

 :  u   u  (pretrained on Krizhevesky et al, 2012.) u Input: image pixels, output: label Noun or Adj; two CNN models. u Generate P(C | I) = f(Wc ∗ I), Wc : CNN’s parameters, I: input image, *: operations of convolution, pooling, activation. u Train via BPTT   …

Slide 21

Slide 21 text

f u 0 r: : 174 : 4 : 0 :41 : 37 : e d → v [ ] 2 20 w2    co 0 e   0 by similarity

Slide 22

Slide 22 text

c ]: hg u a n b . ) a h ) ) ) N X b u N R c [ : R b c [ e: Rb o c [ r(:e Rbo d c [ e Rbod ( [ X EbN? -- i

Slide 23

Slide 23 text

.3* /1=;& u &!(-B JJ u 6  1&8@G!1I71E: u +2E! n-gram 1:A. E.g 4-gram: u 0?7 n 41<4: 4-gram:1X u #<4E1B ",(- B u 1+ C =1C B > 5%1(-$30% u +)DH(B→'GF$9$5%(with beam search) 1      

Slide 24

Slide 24 text

i u rn b a - u g:brn rn i i rn m b b /- b i  b b a b a 2 b i  / b a b a 

Slide 25

Slide 25 text

>:'1;7 --- =, u = 1=H"5<>1 0<>B+6&                       u )## “I”+”” 1)." “8”+””, (1 P(“I”) > P(“I"”) '#A</19 i.e [23G-], [23] @C>:'1% ;7))#<1*4 E E? F!$1= D

Slide 26

Slide 26 text

( m )? u u -: g s d - R s I u N e g A u u n ( ) u r a b

Slide 27

Slide 27 text

74'-61 (charRNN-61) -:< "5/=) ! u #30. 02 ($ 5/) u :<&%?-8' !,  # ";/ u :< 0.8@+  u (61 *->$ u 961

Slide 28

Slide 28 text

CharRNN u RNN model ( CharRNN model u ( ) u

Slide 29

Slide 29 text

No content

Slide 30

Slide 30 text

5";/-@ F u F G.5"7QS u 4P-5"N&,5"E7G 7>O 7K= 7K= D I8H*C 7R'J2 79'J2 1! ?7#%067A: $73 B (3<7 L) L) +M 

Slide 31

Slide 31 text

PU u ;1#I U"HG0 C3\? u H`I C#SK3 UB7-:A[*Q* HIV_ ,L8A[a u 6 C#IUà ' TNIWR< u %2/B7IX] #à$O@JIXY] u #IU'# SK2I>F()I@9.!I + u HV5Z ED u HI^/5&49=IM:

Slide 32

Slide 32 text

b u r g nk u 4 ma- >>  m ma… - >> - >> “ ma- >> - - -ma … - - …. Θ > pi N–gram LM ma|Θ >P( |Θ) >P( |Θ) b

Slide 33

Slide 33 text

N u N-gram N R N u R u

Slide 34

Slide 34 text

k[4p7a_P^+a_M6U@fhVmVe:,,8pQ P* u Q U@o'21F→nP_a_/\PNc u  Q U@J9"A a $→aIba P W4p5j N u fhg3ZSiX. &d 5)_ Pa=?0P u YaQDP;4 YM6U@QDP(C4

Slide 35

Slide 35 text

+$: 1:(' u (=7? /* . u D4+2-3$ 1(%:9 u ,80(;6"-!?B3B(' 2@ & ' u >F(A< u 1'C(GE )5H#

Slide 36

Slide 36 text

8 u 8 010 u  0D4 u < u :?= u  u & u 3628 u General guideline /9<+$*>') $%@*>"F')0A %#C;E.7A(  1+6,D u -68 0D4 8 50  u A/B test, latin square 0! B 

Slide 37

Slide 37 text

u u ) 2 ( 2 ) u 7 ( ) u : ( u ) 0” ”     () : 1

Slide 38

Slide 38 text

" 0*23 u !'*%+)transformer  RNN) u , -*$ /*,# u ,&-(.*- u !3*0 1 

Slide 39

Slide 39 text

Two sentences to conclude u AI AI    u  

Slide 40

Slide 40 text

We are hiring ! u Data Scientist u Data Engineer u Data Analyst u NLP Engineer

Slide 41

Slide 41 text

Thank you QA [email protected]