Papers of the day   All papers

EAR2FACE: DEEP BIOMETRIC MODALITY MAPPING

Comments

Kevin Simler: Well this is astonishing, and not a little uncanny. A neural network can predict what someone’s face looks like, given just an image of their EAR. Frankly a little hard to believe that it works this well....

26 replies, 274 likes


harper 🤯: a minute ago @nikete pointed me to this paper: https://arxiv.org/pdf/2006.01943.pdf EAR2FACE https://t.co/N886AOnAML

20 replies, 188 likes


roadrunner01: Ear2Face: Deep Biometric Modality Mapping pdf: https://arxiv.org/pdf/2006.01943.pdf abs: https://arxiv.org/abs/2006.01943 https://t.co/fRl6OyjD38

2 replies, 128 likes


Shawn Presser: This is the craziest AI project I’ve seen yet. Is this legit? If so, well done.

6 replies, 33 likes


Fun Machine Learning: The CREEPIEST #AI I've ever seen. Creepier than http://thisfootdoesnotexist.com The Ear2Face AI Input a photo of an ear and have it guess what you look like. The scary/creepy part? It's REALLY good https://arxiv.org/pdf/2006.01943.pdf https://t.co/yg6dxYL2Ky

0 replies, 28 likes


Delip Rao: Learning “Ear2Face”. Not sure about the science behind this (building models is not same as scientific validity), but I bet dictators around the would be eager to deploy it in this post-pandemic masked world. #DeepLearning https://arxiv.org/abs/2006.01943 https://t.co/Sm1rcKltl4

5 replies, 23 likes


Paul McKellar: Do you remember the moment you realized we have to cover our ears to protect our identities? https://twitter.com/harper/status/1268358650105643009?s=12

2 replies, 15 likes


Tim Erickson: y’all ready for some doom future technology? turns out ears are very good indicators of face shape, so much so that an AI can roughly predict a face based only on the input of a photo of an ear. full paper here: https://arxiv.org/pdf/2006.01943.pdf https://t.co/4LIBN2wn9M

1 replies, 15 likes


Zauberer Churl: get woke https://t.co/i8iXXL9HS9

1 replies, 9 likes


Reuben Wu: Uh oh

1 replies, 7 likes


Nicholas Frota: inferring faces from EARS!!! (pdf) https://arxiv.org/pdf/2006.01943.pdf https://t.co/BUFDVz4AkX

3 replies, 5 likes


Giorgio Patrini 🛡️👾: 🤨 "our goal is to generate a frontal face image of a subject given his/her ear image as the input"

1 replies, 4 likes


will knight: They do say that the ear is the window to the face.

1 replies, 4 likes


Mathana: I'm pretty ok with pressing the big red STOP button to keep "scientific research" from pursuing phrenological surveillance tech with a stated goal "to generate a frontal face image of a subject given his/her ear image as the input." https://arxiv.org/abs/2006.01943 https://t.co/K448QDjQd7

0 replies, 3 likes


ML Review: Ear2Face: Deep Biometric Modality Mapping 90.9% Rank-10 identification accuracy on the FERET dataset https://arxiv.org/abs/2006.01943 https://t.co/ALwbRb65I7

1 replies, 3 likes


Cem 💎: https://t.co/zZOfvu8Fhv

1 replies, 3 likes


Kosta Derpanis: Should make for an interesting #NeurIPS2020 impact statement.

0 replies, 3 likes


Andrés Romero: "Ear2Face" is a recent method to transform ears images to faces (?) https://arxiv.org/pdf/2006.01943.pdf ... Wait, what? ... Why? https://t.co/7PBjGw8RPs

1 replies, 2 likes


Mike Davies: EAR2FACE: predict a person's face very reliably from just an image of their ear(!). To see how good it is compare generated (every 2/3 column) to real face (3/3 column) in the image. Pandora's box opens w/ uses for: CSI, Doxxing, CCP. paper: https://arxiv.org/pdf/2006.01943.pdf h/t @harper https://t.co/JRINxOgymN

0 replies, 2 likes


Prabhakar Reddy: A neural network can predict what someone’s face looks like, given just an image of their EAR. The team behind this work also built a neural network that can predict someone's face just through their SPEECH PATTERNS. 🤯 https://arxiv.org/pdf/2006.01943.pdf https://t.co/ZJaeVXwLy6

0 replies, 1 likes


RaMi !!!!♿: Now AI can generate a face from a picture of an ear 👂! https://arxiv.org/pdf/2006.01943.pdf https://t.co/aC8tyuTu9G

1 replies, 1 likes


Roberta: @rafaeu https://arxiv.org/pdf/2006.01943.pdf

1 replies, 1 likes


Content

Found on Jun 04 2020 at https://arxiv.org/pdf/2006.01943.pdf

PDF content of a computer science paper: EAR2FACE: DEEP BIOMETRIC MODALITY MAPPING