Huang Dong’s Blog, email: huangdongxy@hotmail.com

January 19, 2009

experiments based on yaleB

Filed under: pattern recognition, Tech — donghuang @ 12:40 am

tasks:

  1. exp1: repeat the experiments in paper:
    1. illumination extrapolation: trainSet: frontal pose P00, 7 illumination at 12degree: 7 x 1pose x 10perseon = 70; testSet1: 38(illumination between 12 degree and 77 degree) x 1pose(P00) x 10person = 380
    2. pose and illumination extrapolation: trainSet: as above; testSet2: 380 + 45(illumination < 77degree) x 8poses x 10person = 3980
  2. exp2: pose extrapolation.  pick one pose for training, the other poses for test. trainSet: as above; testSet: 7illumination x 8pose x 10person=560
  3. exp3: illumination extrapolation. pick the 7 frontal illumination for training, the others for test. TrainSet: as above; testSet: 38illunimation x 1pose(P00) x 10person =380
  4. exp4: trainSet: 8 frontal illumination x 9pose x 10person, testSet: 38 illumination x 9pose x 10person
  5. exp5: trainSet: random pick 20 illumination x 90pose x 10person, testSet: 26illumination x 9pose x10person

what I have done:

  • extractecd the tar.gz files into 90 folders
  • load the images in the 90 folders and save to yaleB.mat
  • align the face images: for frontal pose, affine transform is used since there are 3 pairs of control points; for other poses, images are shifted to have the face centered at the center of the image
    • tform = cp2tform(input_points,base_points,’affine’); %cp2tform(input_points, base_points, transformtype)
    • B = imtransform(A,tform,interp,’XData’,[182,457],’YData’,[108,404],’XYScale’,1); %interp=’bicubic’
    • visually compare the aligned face images
      • found bizarre images for person 2 pose 00, further inspection needed.
      • due to affine transformation for the frontal pose, the frontal pose and other poses of the same person look distorted for some persons. However, train images are only from the frontal pose for exp1. A better way to register face images may be to just translate the images to align the face centers. This way train and test images may be more consistent.
  • histogram equalization: each face image is divided into 2 halfs (left, right), and histeq is applied on each half image.
  • partition the database into training set and 2 test sets
    • trainSet: frontal pose P00, 7 illumination at 12degree: 7 x 1pose x 10perseon = 70
    • testSet1: 38(illumination between 12 degree and 77 degree) x 1pose(P00) x 10person = 380
    • testSet2: 380 + 45(illumination < 77degree) x 8poses x 10person = 3980
  • train the system on trainSet, obtain the eigenface, fisherface, reconstructed images, etc.
  • test the trained system using the 2 test sets.
    • 3 subsets for illumination: 12degree, 25degree, 50degree,77degree
    • 3 subsets in pose: frontal (Pose 1), 12degree (pose 2~6), 24degree (pose 7~8)
  • compare the results to the published results

 works to be done:

  • get detailed classification accuracies for each subset.
    • get results for rmcld first and compare with the published result.
  • perform exp 4
    • compare PCA, RMLD, RCLD, RBCLD

January 13, 2009

十二星座之念念不忘的爱情

Filed under: life — donghuang @ 1:53 am

十二星座之有一段情一直住在心里

到底什么样的恋情,会让他们念念不忘,有如篆刻在心?

火象星座:太美好的恋情(白羊座、狮子座、射手座)

你爱上了一个孩子,那时候你同样是个孩子,你们一起经历生命中第一次的相思与熬煎,第一次的相恋与反叛,发自肺腑,情真意切,可以为爱走天涯,掺不得半点儿私心杂念。

多少年后,无论陪伴在你身边的那一个是谁,你都会在某时某刻,想起当年。所有情景历历在目,历久弥新,你可能会笑,也可能哭泣,你会惊奇地发现,原来自己的内心并不如想象中凌乱杂沓,偶尔那些干净的美好会光临,虽然回不去。

那一场爱情,像一朵花,绽放得急切,刚刚娇艳,转眼就雨打风吹了。留下唯美的感动,缺憾的动容,却像是把美丽一片片张扬地撕碎了给人看。

之后,你会明白,生死爱恨敌不过时空变错。也会明白,美好,并不是奢华,而是一种来自心灵的幸福。那时候,或许你并没有后来的物质,也正如后来永远失去了那时的纯真一样。

土象星座:没有结果的恋情(金牛座、处女座、摩羯座)

有段恋情,你付出了许多,全心全意,以为可以归宿圆满,却不想,自己甚至连一个凄美的结局都没有等到。爱的终结,像一场闹剧,如果可以真的从脑海抹去,现实便不会残留一丝痕迹;分离的一刻,发生时犹如晴天霹雳,来不及反应已经灭顶。

于是,你会惦念着,情不自禁地问自己,到底出了什么问题,时而你会原谅对方,体念着他的苦衷,时而你又充满了怨恨,在怀念中延展。谁对谁错,没有理由,只有借口。

有些人,你拼命地想为他(她)改变,最后仍然站在原地;有些人你从未留意,却在不知不觉中,为他(她)改变了。可改变后,他(她)又不再停留,只有你独自纠结太没有结果的恋情。

旧的总要被新的代替,就像雨果说的,一个杀死另一个。他(她)杀死的,到底是一场爱情,一份纯真,还是一个你。那时,你并不知道,在路的尽头,等着你的那个人原来不是他(她)。

水相星座:太痛楚的恋情(巨蟹座、天蝎座、双鱼座)

亦舒的小说,写一个暗恋男人的女子对另一个承欢的女子说:“你放过他吧。”那个承欢的女人恐怕也极爱他,所以才会笑着说:“什么叫放过他,根本相爱的人互不放过。”痛,缘于对伤口的摩擦,你惦念,便表示你一直都未停止对伤口的折磨。

时过境迁,或许有两种可能,你开始害怕,便缩到一个壳里;你肆无忌惮,挥霍着所有爱的可能。你必须寻找一个有能力抚平伤口的人,不然便是无尽的痛楚,他或许早已放了你,你却没有那么洒脱。

或者,你并不是要一个答案,一个理由,而是爱直至成伤后,悲观的情绪无限放大,涌出。你不甘就这样放下,却也不敢再去轰轰烈烈、你死我活,于是,积压成抹不去的回忆。

其实,当日子累计到可以看透沧桑,你会发现,原来自己曾经珍视的很多东西其实都是虚无,真正无价的却被忽略。突然发觉,伤痛过后是谁会陪伴自己,拼杀得伤痕累累时,又是谁始终不离开你?

风象星座:太孤单的恋情(双子座、天秤座、水瓶座)

如果当年,他说喜欢你的眼睛,那么你会买最好的眼霜去滋养维持,想象会再见到他时,依然能闪烁出美好的光芒。其实,你并不知道他这句赞美的意义是什么,或许他当时也只是随口说说,或者搬出了哪个课文上的句子,但你会当真。

也许,人有三样东西是没法隐瞒的,咳嗽、贫穷和爱。越隐瞒,越欲盖弥彰。可是很多时候你仍想隐瞒,出于天生的不安,环境的约束,或者其他理由,然而,无法诉说的爱情,却比能告白的爱情来得更殷切。这殷切会一直持续,被幻想完善,坚不可摧。

肯定会有人指责,说不真实。甚至,惦念得严重了,你也会想,是自己错误的美化而已。可什么是真实?历史都能篡改,何况只是你心中的一个传说。

你从未想过,匆匆的交汇,竟让自己孤单地挂念一生。这种甜蜜,这种苦痛,没人赞成,没人了解,也可能根本没有人知道。一切,不知是劫是缘。

January 12, 2009

Protected: Efficient BackProp

Filed under: Neural networks, Tech — donghuang @ 11:54 pm

This content is password protected. To view it please enter your password below:

January 9, 2009

oth s4

Filed under: Uncategorized — donghuang @ 10:25 am

there are moments in our lives when we find ourselves at crossroads. The choices we make in those moments can define the rest of our days. Of course, when faced with the unknown, most of us prefer to turn around and go back.

January 8, 2009

oth s1

Filed under: Language Study, life — donghuang @ 5:25 pm

s1 e02, 18min: you ever wake up from a really good dream and try to get back to sleep? or you got the flu and you promise yourself that you’d appreciate normal so much more if you could just get back to it?

January 7, 2009

Protected: The Yale Face Database B

Filed under: pattern recognition — Tags: , , , — donghuang @ 5:46 pm

This content is password protected. To view it please enter your password below:

Create a free website or blog at WordPress.com.