Julien Valentin

Senior Research Scientist at Google. Founding team member of PerceptiveIO. Oxford PhD.

[Julien Valentin @JPCValentin]

https://t.co/sLYjwfuU1m

[Andrea Tagliasacchi 🇨🇦→✈️→🇺🇲 (@CVPR) @taiyasaki]

With #COVID19 many research internship will be cancelled... so why not participate in a Google Summer of Code #GSoC instead, be exposed to cutting edge research, and make some extra $$?

#TensorFlowGraphics have several projects posted here: https://t.co/bvEajxyGq2

[_TFGraphics_ @_TFGraphics_]

Vote for state of the art baselines we will implement in TensorFlow Graphics during the Google Summer of Code!

https://t.co/1ePKzRYTCI

[Julien Valentin @JPCValentin]

@zzznah @GoogleColab Very cool! ;)

[Julien Valentin @JPCValentin]

Get Ready for the ARCore Depth API

A great story of research getting into a product reaching millions of users!

https://t.co/t54mxvpDKt

[Srinath Sridhar @drsrinathsridha]

If you are at #NeurIPS2019 , don't miss our paper on sparse multi-view (1 or more) object reconstruction. **Wed Dec 11th 10:45 AM -- 12:45 PM @ East Exhibition Hall B + C #96** Joint work with: @davrempe @JPCValentin @_sofien_ Leo Guibas. @NeurIPSConf

[Matthias Niessner @MattNiessner]

ViewAL: active learning with viewpoint entropy! https://t.co/HrVIXGESIn

Only 7% training data, achieves 95% network performance!

Simple idea:
If network makes inconsistent predictions across views, we need more training data for these cases :)

#DeepLearning-now-needs-less-data

[Julien Valentin @JPCValentin]

@zkajdan Keras is indeed the way to go.

Best.

[Julien Valentin @JPCValentin]

@zkajdan I unfortunately did not plan what you described for the next release. I would nevertheless be happy to have the discussion with you if you would be interested to contribute.

Best.

[_TFGraphics_ @_TFGraphics_]

Find the #SIGGRAPH2019 course notes on Differentiable Graphics with TensorFlow 2.0 at https://t.co/bn9xp2XyPI

@JPCValentin @_sofien_ @random_forests @shancarter @DynamicWebPaige @zzznah
@martin_wicke

[Julien Valentin @JPCValentin]

@drsrinathsridha The demo scene is quite fun.

Some ML people should generate an hours long video generated from a recurrent net to show them what we can do :P

[Julien Valentin @JPCValentin]

Good morning #SIGGRAPH2019! Join us between 9 and noon in room 408AB for a course on Differentiable Graphics with TensorFlow 2.0. See you there!

@_TFGraphics_ @TensorFlow @_sofien_ @random_forests @shancarter @DynamicWebPaige @martin_wicke

[Julien Valentin @JPCValentin]

@JamieDJS @Bombardier Congratulation @JamieDJS ; much deserved ;)

[Paul Debevec @debfx]

Our Deep View Synthesis CVPR 2019 paper is now online https://t.co/PYGENZA3ma , with examples of Light Field Video https://t.co/4X8vkAqJgm and an interactive multi-plane image (MPI) viewer to try the view interpolation technique for yourself https://t.co/0aplX6MinP

[Srinath Sridhar @drsrinathsridha]

How can we estimate the 6D pose and size of novel object instances that we have never seen before? Come find how we do category-level 6D pose and size estimation at our #CVPR2019 oral today (~15:00, Promenade Ballroom, Poster #181). Code/data available! https://t.co/XtH18ftgdg

[Julien Valentin @JPCValentin]

Today I will be talking about @_TFGraphics_ at the #cvpr2019 3D-WiDGET workshop. See you there! https://t.co/nHR73IUgBg

[Julien Valentin @JPCValentin]

@MattNiessner Well deserved. Congrats 😉

[Julien Valentin @JPCValentin]

@DynamicWebPaige @martin_wicke Ahaha nice one :)

[Julien Valentin @JPCValentin]

@H___Rearden The Trimesh library seems to handle .stl; did you give that a try?

[Julien Valentin @JPCValentin]

@KavitaKrish21 @_sofien_ @podlipensky @kiamada @AvneeshSud I have to confess that I never used TensorFlow.js so can't confirm at this time =(

[Julien Valentin @JPCValentin]

TensorFlow Graphics is out! This new library brings 3D and graphics functionalities to #TensorFlow!

@_sofien_ Cem Keskin @podlipensky @kiamada @AvneeshSud

#io19 #DeepLearning #GoogleARVR #GoogleAI

https://t.co/JIgsoA6k7I