Paper Reviews Call 002 -- FaceNet: A Unified Embedding for Face Recognition and Clustering

Поделиться
HTML-код
  • Опубликовано: 5 ноя 2024

Комментарии • 47

  • @alexn2566
    @alexn2566 5 лет назад +5

    Awesome explanation of the Inception architecture: "If you don`t know what kernel to use just use them all and let the network figure out what to use."

    • @alexn2566
      @alexn2566 3 года назад

      @Lyle Rocco is this marketing? :))

  • @deeps-n5y
    @deeps-n5y 3 месяца назад

    This was so cool! Thank you Tess for this wonderful explanation!

  • @jm10oct
    @jm10oct 3 года назад +3

    THAT was a great great show right here. Thank you so much. that was very helpful. Great Work!

  • @zerotoone4565
    @zerotoone4565 4 года назад +4

    Great talk, thanks for uploading this!

  • @edv9775
    @edv9775 3 года назад

    Thank you for the upload! It helped a lot in understanding FaceNet.

  • @adwaitdhamane8880
    @adwaitdhamane8880 2 года назад

    I was just trying to read this paper. I was not able to understand it very much but this gave me a very good overall understanding of the paper.

  • @harrrymok
    @harrrymok 4 года назад +1

    thanks for the nice review, when i load the facenet model by tf.keras.models.load_model('facenet_keras.h5') with tensorflow 2.2, it keeps raising error (ValueError: bad marshal data (unknown type code), may i know is there any idea to mitigate it?

  • @x9466x
    @x9466x 9 месяцев назад +1

    Thanks for sharing

  • @catalinavram3187
    @catalinavram3187 4 года назад

    This is a great video and a great Channel I just found!!

  • @krupalshah1487
    @krupalshah1487 3 года назад

    Thank you for sharing it ! It's amazing...

  • @khadijakhaldi6468
    @khadijakhaldi6468 5 лет назад +1

    Hello, Thank you so much for the explanation. Please caw you share with us the notebook ? Thank you!

  • @waleedaiad3411
    @waleedaiad3411 2 года назад

    How can i khow how many hidden layer used in facenet?

  • @tatendaemmamatika4131
    @tatendaemmamatika4131 2 года назад

    At 13:21 - With DALL·E we can create beach jam! 😂😂😂

  • @asharkhan6714
    @asharkhan6714 5 лет назад

    I love you for uploading this!

    • @GabitTolendiyev
      @GabitTolendiyev 5 лет назад +1

      Did you try this? Didn't you have any problem?

    • @asharkhan6714
      @asharkhan6714 5 лет назад

      @@GabitTolendiyev Yeah I tried it using opencv and a caffe model and an SVM for classification. It works great so far. I'm creating a GUI desktop app for this and will be uploading it to my github so you can take a look at github.com/ashar-7

  • @nidhirbhavsar3917
    @nidhirbhavsar3917 4 года назад +1

    Good job

  • @masoncusack
    @masoncusack 5 лет назад +3

    This if fantastic. 😍

  • @MultiSuperRita
    @MultiSuperRita 5 лет назад

    Terrific! thank you for making it public.

  • @abhinayoja272
    @abhinayoja272 5 лет назад

    Great video. Where can I find the notebook

    • @machinelearningdojo
      @machinelearningdojo  5 лет назад +1

      I'll ask Tess to publish it!

    • @abhinayoja272
      @abhinayoja272 5 лет назад

      @@machinelearningdojo thanks!
      that would be phenomenal

    • @abhinayoja272
      @abhinayoja272 5 лет назад

      @@machinelearningdojo sorry to bother you once again,but could you please share the notebook

    • @siddheshghadi9207
      @siddheshghadi9207 5 лет назад

      Plzz share the notebook

    • @machinelearningdojo
      @machinelearningdojo  5 лет назад

      @@siddheshghadi9207 github.com/TessFerrandez/research-papers/blob/master/facenet/FaceNet.ipynb

  • @GabitTolendiyev
    @GabitTolendiyev 5 лет назад

    I have a problem,
    RuntimeError: Unable to open models/landmarks.dat
    Could you help me please?

    • @tessferrandez7773
      @tessferrandez7773 5 лет назад +1

      you either need to train your own, or find/download a pre-trained off the internet... I am unfortunately not allowed to distribute it

    • @GabitTolendiyev
      @GabitTolendiyev 5 лет назад

      @@tessferrandez7773 what I need to train? Lots of data? Thanks

    • @tessferrandez7773
      @tessferrandez7773 5 лет назад +1

      Gabit Tolendiyev if you want to train your own then yes, but you can use a pre-trained model and just generate embeddings from single or few photos per person

    • @GabitTolendiyev
      @GabitTolendiyev 5 лет назад

      @@tessferrandez7773 thank you!

    • @DeepakKumar-uz4xy
      @DeepakKumar-uz4xy 4 года назад +1

      @@tessferrandez7773 mam how to generate embeddings from images

  • @keenlearner8751
    @keenlearner8751 5 лет назад

    Is this IEEE Transaction
    The paper

  • @wilgeraleman7940
    @wilgeraleman7940 5 лет назад

    could you provide the slides of the presentation?

    • @tessferrandez7773
      @tessferrandez7773 5 лет назад

      Hi Wilger, I have uploaded them here www.slideshare.net/TessFerrandez/facenet-paper-review

  • @DeepakKumar-uz4xy
    @DeepakKumar-uz4xy 4 года назад

    can we use cosine in facennet

    • @lynnwilliam
      @lynnwilliam Год назад

      In theory yeah, but I have a DB of 70,000 faces. And searching for 1 face can find 1000 or so results. And this is being conservative.