Arpan Chakraborty (@runoffthemill) profile picture

Autonomous Driving @ Ridecell | Ex Instructor, Udacity | Ex Adjunct Lecturer, Georgia Tech

Joined October 2014
Photo by Arpan Chakraborty (@runoffthemill)Photo by Arpan Chakraborty (@runoffthemill)
Arpan Chakraborty retweeted
More than 100 students at a high school in Huntington, W.Va., staged a walkout to protest a Christian assembly they were forced to attend during school last week, chanting "separate the church and state" and "my faith, my choice."
85
302
90
1,692
Arpan Chakraborty retweeted
We hosted the amazing @garrytan at @Cruise to share his wisdom on building world-changing companies. As you can imagine, it was epic. “I would say a quarter of GDP is today being properly serviced by software engineers and good software.” Transportation included!
1
6
1
75
Arpan Chakraborty retweeted
I'll speak about deep reinforcement learning and policy gradients tonight at 7 pm (EST) in Manning at Twitch. No background in ML needed. Don't miss it!
So how does deep reinforcement learning actually work? Join @luis_likes_math on Twitch today, 7PM EDT for a friendly introduction to deep reinforcement learning twitch.tv/manningpublication… #ReinforcementLearning #MachineLearning #DeepLearning
3
12
Arpan Chakraborty retweeted
Today Ridecell announces a breakthrough in harnessing the power of IoT–Ridecell #Fleet #IoT #Automation. The platform complements the existing # #fleetmanagement and vehicle #telematics solutions and unifies data from these systems to automate tasks. https://teapawt.org/t.co/Ehw5k6KI5l
9
13
Driving scenarios provide a robust framework for testing & validation of #autonomousdriving systems. Learn about the fundamental elements & spatiotemporal relationships that characterize scenarios, & build an expressive language for representing them. lnkd.in/gis8U4n
Autonomous Vehicles have to share the road with all kinds of drivers. Modeling their behaviors (and reproducing them in simulation) is essential for training our systems to respond appropriately. #AutonomousVehicles #TrafficModeling #Simulation
AV and ADAS teams need to use data and simulation more effectively than ever in the current pandemic. Here is a glimpse of how we create human-like driving behaviors in simulation platforms, using real-world data distributions auro.ai/blog/2020/10/creatin…
1
What does it take to learn effective driving policies for #AutonomousVehicles in simulation? 1. Real-world scenarios 2. Data-driven traffic actors 3. Modular learning agents 4. Loads of naturalistic driving data! Check out our blog -auro.ai/blog/2020/05/acceler… #ReinforcementLearning
1
5
7
For all the #ReinforcementLearning Enthusiasts out there, we are sharing a beginners guide for training Deep RL agents for the task of #autonomous_driving. Check out our blog and try training your own RL agents to drive in #Unity with the code we shared. auro.ai/blog/2020/05/learnin…
4
2
16
Here is a sample of how we automated the process of scenario mining and multiple scenario generation from a single template (simulated in #CARLA). This is a must for both system validation and learning complex driving behaviors at an accelerated pace. #AutonomousVehicles #L4
4
9
Arpan Chakraborty retweeted
A new breakthrough reduces the percentage of cobalt in lithium-ion batteries. Now that may not sound like much to most readers, but that reduction lowers the cost of #EVs, boosts their range, and has big implications for the future of #electricvehicles. marker.medium.com/a-new-batt…
1
3
Arpan Chakraborty retweeted
COVID-19 Should Be Wake-Up Call for Robotics Research ri.cmu.edu/covid-19-should-b…
1
1
6
Arpan Chakraborty retweeted
Couldn't be more excited about @udacity 's willingness to offer a month of our classes for free. This is an unprecedented situation, and none of us knows where it will lead us. I sincerely hope we will emerge strong and united.
#COVID19 has changed our lives. In a scenario when education and workforce are experiencing some of the biggest challenges, learning new #skills should be made more accessible. Hence, we're excited to announce one free month on 40+ Nanodegree programs. https://teapawt.org/t.co/noRYuDF4Pm
15
28
6
142
Arpan Chakraborty retweeted
At home and interested in learning some NLP? Check out my two series of videos on Latent Dirichlet Allocation! Video 1: Blueprint of LDA youtube.com/watch?v=T05t-SqK… Video 2: Training LDA using Gibbs sampling youtube.com/watch?v=BaM1uiCp…
6
14
1
54
Wondering how to run @ApolloPlatform with @carlasimulator? Easy - write a bridge and convert Carla's #OpenDRIVE maps to Apollo. Well, there's more to it! Read on: https://teapawt.org/t.co/Chjd1ARjga
This quoted post is unavailable.
Arpan Chakraborty retweeted
Meet Hammerhead, the patent-pending, vehicle-agnostic #autonomous sensor unit from @Auro_AI, the Autonomous Driving Division of Ridecell, that's bringing us one step closer to achieving real-world, near-term applications of L4 #autonomy for shared #fleets. ridecell.com/blog/a-modular-…
6
9
Excited to launch Ichigo, our next-generation #autonomousvehicle with new #selfdriving capabilities for public roads, an upgraded sensor suite and a sleek rooftop autonomy sensor box. Come check it out at the Computer History Museum waj.org/2019-silicon-valley-…
9
1
11
Auro by @ridecell has open-sourced a message communication bridge between Apollo, #Baidu's opensource AV stack and ROS. For the first time ever, it will be possible to use the best features of both @ApolloPlatform and @rosorg together. Read more: auro.ai/blog/2019/08/using-o…
7
6
Enlightening and witty as always - @SebastianThrun thanks for sharing your wisdom, and showcasing @Udacity student projects in your talk! #DriveWorldESC
4
Watching my buddy @SinaBahram talk about #a11y and inclusive design. Refreshing!
NEW VIDEO: In his keynote address delivered at the 2019 Tessitura Learning & Community Conference, @SinaBahram challenges us to use the principles of inclusive design to open up our buildings, content, programming & experiences. View here: https://teapawt.org/t.co/MPojAjpxYE #TLCC2019 #a11y
2