M. Alex O. Vasilescu (@alexvasilescu) 's Twitter Profile
M. Alex O. Vasilescu

@alexvasilescu

Developing a #causal #tensor framework: TensorFaces, Human Motion Signatures| @MIT & @UofT alumna| @UCLA Assoc Dir #ComputerVision & #Graphics Lab| TensorVision

ID: 1294920027175452676

linkhttp://web.cs.ucla.edu/~maov/ calendar_today16-08-2020 08:53:17

945 Tweet

608 Takipçi

752 Takip Edilen

Walter Scheirer (@wjscheirer) 's Twitter Profile Photo

The November issue of the PAMI-TC newsletter is out: tinyurl.com/bn56czrm Some items of note: 1. At the urging of some members of the community, the CVPR/ICCV Steering Committee has reviewed several of the new author policies for #CVPR2025 and endorsed them. 2. arXiv.org is

The November issue of the PAMI-TC newsletter is out: tinyurl.com/bn56czrm

Some items of note:

1. At the urging of some members of the community, the CVPR/ICCV Steering Committee has reviewed several of the new author policies for #CVPR2025 and endorsed them.

2. <a href="/arxiv/">arXiv.org</a> is
Federico Becattini (@fede_becat) 's Twitter Profile Photo

#CVPR2025 #CVPR2025 a note: if you set deadlines at 2AM people might see the date and assume they have more time. Authors should be careful, but it can happen. Better to set it a couple of hours before, say 23.59PM. Even better, use AoE timezone so nobody gets confused :)

<a href="/CVPR/">#CVPR2025</a> #CVPR2025 a note: if you set deadlines at 2AM people might see the date and assume they have more time. Authors should be careful, but it can happen. Better to set it a couple of hours before, say 23.59PM. Even better, use AoE timezone so nobody gets confused :)
M. Alex O. Vasilescu (@alextensor) 's Twitter Profile Photo

Fight the matrix, use #tensor algebra and perform #causaldeeplearning! arxiv.org/pdf/2301.00314 #causality #causalTwitter #machinelearning #computervision #artificial_intelligence

Fight the matrix, use #tensor algebra and perform #causaldeeplearning!

arxiv.org/pdf/2301.00314

#causality #causalTwitter  #machinelearning #computervision #artificial_intelligence
Percy Liang (@percyliang) 's Twitter Profile Photo

While we celebrate DeepSeek 's release of open-weight models that we can all play with at home, just a friendly reminder that they are not *open-source*; there’s no training / data processing code, and hardly any information about the data.

M. Alex O. Vasilescu (@alextensor) 's Twitter Profile Photo

Interesting exchange -- Is #AI a tool or an agent with its own goals and desires? --- @DarioAmodeig ,Andrew Ng,Yoshua Bengio and Yejin Choi at #WEF25. Andrew Ng : "There are two philosophies on AI that are being represented here... I tend to view AI as a tool and it is

M. Alex O. Vasilescu (@alextensor) 's Twitter Profile Photo

The cost of training and generating OpenAI synthetic data ought to be amortized into the $6M spent by DeepSeek. #AI #LLM ---- venturebeat.com/ai/calm-down-d… via VentureBeat "....There’s no denying that DeepSeek-R1’s cost-effectiveness is a significant achievement. But let’s not

M. Alex O. Vasilescu (@alextensor) 's Twitter Profile Photo

What was the point you were trying to make, Geoffrey Hinton? #AI youtu.be/vxkBE23zDmQ?si… While replacing a single neuron with nanotechnology may preserve #consciousness, it does not necessarily follow that replacing all neurons will do the same or that the resulting nanotech

M. Alex O. Vasilescu (@alextensor) 's Twitter Profile Photo

TensorFaces is a 4th order tensor where there are a total of I_pixels x I_illuminations x I_views x I_people). A tensor T is a multilinear mapping from a set of domain vector spaces to a range vector space, T : {C^ I_illuminations × C^I_views × C^I_people } → C ^I_pixels .

TensorFaces is a 4th order tensor where there are a total of I_pixels x I_illuminations x I_views x I_people).

A tensor T is a multilinear mapping from a set of domain vector spaces to a range vector space, 
T : {C^ I_illuminations × C^I_views × C^I_people } → C ^I_pixels .
#ICCP2025 (@iccp_conference) 's Twitter Profile Photo

📢Attention researchers in computational imaging! The Call for Papers page for #ICCP2025 is live. We invite high-quality submissions on theory, sensors, optics, algorithms, and much more 📷. Submit your work by April 9, 2025. #ICCP2025 #CallForPapers

📢Attention researchers in computational imaging! The Call for Papers page for #ICCP2025 is live. We invite high-quality submissions on theory, sensors, optics, algorithms, and much more 📷. Submit your work by April 9, 2025. #ICCP2025 #CallForPapers
Walter Scheirer (@wjscheirer) 's Twitter Profile Photo

The March issue of the PAMI-TC newsletter is out: tinyurl.com/56my336j Perfect reading if you need a break from editing your #ICCV2025 paper. This edition has a bunch of calls for #CVPR2025, #ICCV2025, and #ICCP2025. There are plenty of open opportunities to participate in

The March issue of the PAMI-TC newsletter is out: tinyurl.com/56my336j

Perfect reading if you need a break from editing your #ICCV2025 paper. This edition has a bunch of calls for #CVPR2025, #ICCV2025, and #ICCP2025. There are plenty of open opportunities to participate in
M. Alex O. Vasilescu (@alextensor) 's Twitter Profile Photo

In 1956, John McCarthy, Marvin Minsky (my MIT CSAIL academic advisor), N. Rochester, and Claude Shannon set out to precisely define learning and intelligence so that a machine could simulate them. They aimed to determine how a machine could use language, form abstractions and

In 1956, John McCarthy, Marvin Minsky (my <a href="/MIT_CSAIL/">MIT CSAIL</a> academic advisor), N. Rochester, and Claude Shannon set out to precisely define learning and intelligence so that a machine could simulate them.  They aimed to determine how a machine could use language, form abstractions and
M. Alex O. Vasilescu (@alextensor) 's Twitter Profile Photo

2024: Causal Deep Learning lnkd.in/dpS38YhN 2020: Causal eXplanations with Block Tensor Factorization lnkd.in/dfvnSbex 2019: Hierarchical Tensor Factor Analysis lnkd.in/gMaYrQz 2007/2011: Inverse Causal Inference / Multilinear Projection

M. Alex O. Vasilescu (@alextensor) 's Twitter Profile Photo

Mimicking emotional behaviors is not the same as having or experiencing emotions. It is concerning to see Hinton anthropomorphizing machines and promoting magical thinking. This is dangerous. It invites the public to misunderstand technical malfunctions as personality