Christopher Xie

Graduate Student at UW, studying ML, Computer Vision, Robots!

[Reflection Robotics @ReflectionR0b0t]

Happy Pi day!
To celebrate the π Day, we challenged our robot to type the first 12 digits of $\pi$ with a gripper 2x the size of buttons. On top of that, the calculator is constantly moved to force an error but we failed.

#robot

[Xiaoyang Wu @XiaoyangWu_]

📢Sonata: Self-Supervised Learning of Reliable Point Representations📢

Meet Sonata, our"3D-DINO" pre-trained with Point Transformer V3, accepted at #CVPR2025!

🌍: https://t.co/jIte27gVmW
📦: https://t.co/qqEOChCZ86
🚀: https://t.co/GQ71nMwIVu

🔹Semantic-aware and spatial reasoning representations learned with no label;
🔹3x linear probing accuracy (from 21.8% to 72.5%) on ScanNet;
🔹2x data efficiency performance with only 1% of the data compared to previous approaches;
🔹As always, establish new SOTA results across indoor and outdoor 3D perception tasks.

Our author team: @HengshuangZhao, @jstraub6, @rapideRobot, @ddetone, @NinjaDuncan, @TianweiS, @Christopher_Xie, @NanYang719.

[Project Aria @Meta @meta_aria]

Check out our extension of SceneScript to human-in-the-loop local corrections!

Our method leverages infilling techniques from NLP to refine a 3D scene in a "one-click fix" workflow, enabling more accurate modeling of complex layouts.

📰https://t.co/exbbLM8yFP
🔗https://t.co/TZIzmFmDWg

[Christopher Xie @Christopher_Xie]

@chris_j_paxton This is big news! Good luck on the next adventure!

[AI at Meta @AIatMeta]

Today we’re introducing SceneScript, a novel method for reconstructing environments and representing the layout of physical spaces from @RealityLabs Research.

Details ➡️ https://t.co/1zErU9WT9t

SceneScript is able to directly infer a room’s geometry using end-to-end machine learning and represent it using language. Compared to previous approaches, this results in representations of physical scenes that are compact, complete, interpretable and extensible.

[Edward Miller @TweetEdMiller]

Interested in accelerating real-world scene understanding for AR and AI technologies?

Introducing ‘Aria Synthetic Environments’, a large-scale dataset consisting of procedurally-generated simulated scenes for accelerating indoor environment understanding research.

#CVPR2023
1/7

[Christopher Xie @Christopher_Xie]

@bauzavillalonga @DeepMind Congrats! And welcome to London! We should grab lunch sometime, I'm literally right across the street from you :D

[Christopher Xie @Christopher_Xie]

@SamuelAinsworth @UW @uwcse @uw_wail Yes I am! Let me know whenever you're in town!

[Christopher Xie @Christopher_Xie]

Check out our recent work: FiG-NeRF: Figure-Ground Neural Radiance Fields for 3D Object Category Modelling!

TLDR; we adapt NeRF to learn category models from in-the-wild images.

Website: https://t.co/erFQ6LheR9
Paper: https://t.co/tgRpaOiFGx
Video: https://t.co/08xMhlJHB3

[Christopher Xie @Christopher_Xie]

@michellearning @willie_agnew @samarth_robo Sounds good! @willie_agnew I'll let you coordinate, thanks!

[Christopher Xie @Christopher_Xie]

@willie_agnew @michellearning @samarth_robo Yeah, meeting up and chatting would be great! Sounds like a lot of fun. @michellearning are you in Seattle for the time being?

[Christopher Xie @Christopher_Xie]

@michellearning This is awesome, thanks for the recommendation. Do you have any recommendations to study up on some of the basics of this stuff? My knowledge is severely lacking when it comes to equations of motion, dynamics, etc., and I am looking to improve my knowledge of it.

[Christopher Xie @Christopher_Xie]

@vivjay30 @NeurIPSConf Wow, these results look and sound great!

[sim2real @sim2realAIorg]

Learning RGB-D Feature Embeddings for Unseen Object Instance Segmentation

https://t.co/tfyBDuJUVf

[Christopher Xie @Christopher_Xie]

@Adithya_Murali_ @LerrelPinto @a__mousavian @clembow @chris_j_paxton Wow congrats!!!

[Christopher Xie @Christopher_Xie]

@SimonShaoleiDu @UW @uwcse Congrats Simon! Welcome to UW!

[Christopher Xie @Christopher_Xie]

Check out our work recently published at CoRL 2019!

[sim2real @sim2realAIorg]

The Best of Both Modes: Separately Leveraging RGB and Depth for Unseen Object Instance Segmentation

https://t.co/bW0KcU4MRT

wepage: https://t.co/xAym1Wyqzf