More from All
How can we use language supervision to learn better visual representations for robotics?
Introducing Voltron: Language-Driven Representation Learning for Robotics!
Paper: https://t.co/gIsRPtSjKz
Models: https://t.co/NOB3cpATYG
Evaluation: https://t.co/aOzQu95J8z
🧵👇(1 / 12)
Videos of humans performing everyday tasks (Something-Something-v2, Ego4D) offer a rich and diverse resource for learning representations for robotic manipulation.
Yet, an underused part of these datasets are the rich, natural language annotations accompanying each video. (2/12)
The Voltron framework offers a simple way to use language supervision to shape representation learning, building off of prior work in representations for robotics like MVP (https://t.co/Pb0mk9hb4i) and R3M (https://t.co/o2Fkc3fP0e).
The secret is *balance* (3/12)
Starting with a masked autoencoder over frames from these video clips, make a choice:
1) Condition on language and improve our ability to reconstruct the scene.
2) Generate language given the visual representation and improve our ability to describe what's happening. (4/12)
By trading off *conditioning* and *generation* we show that we can learn 1) better representations than prior methods, and 2) explicitly shape the balance of low and high-level features captured.
Why is the ability to shape this balance important? (5/12)
Introducing Voltron: Language-Driven Representation Learning for Robotics!
Paper: https://t.co/gIsRPtSjKz
Models: https://t.co/NOB3cpATYG
Evaluation: https://t.co/aOzQu95J8z
🧵👇(1 / 12)

Videos of humans performing everyday tasks (Something-Something-v2, Ego4D) offer a rich and diverse resource for learning representations for robotic manipulation.
Yet, an underused part of these datasets are the rich, natural language annotations accompanying each video. (2/12)
The Voltron framework offers a simple way to use language supervision to shape representation learning, building off of prior work in representations for robotics like MVP (https://t.co/Pb0mk9hb4i) and R3M (https://t.co/o2Fkc3fP0e).
The secret is *balance* (3/12)
Starting with a masked autoencoder over frames from these video clips, make a choice:
1) Condition on language and improve our ability to reconstruct the scene.
2) Generate language given the visual representation and improve our ability to describe what's happening. (4/12)
By trading off *conditioning* and *generation* we show that we can learn 1) better representations than prior methods, and 2) explicitly shape the balance of low and high-level features captured.
Why is the ability to shape this balance important? (5/12)
You May Also Like
So friends here is the thread on the recommended pathway for new entrants in the stock market.
Here I will share what I believe are essentials for anybody who is interested in stock markets and the resources to learn them, its from my experience and by no means exhaustive..
First the very basic : The Dow theory, Everybody must have basic understanding of it and must learn to observe High Highs, Higher Lows, Lower Highs and Lowers lows on charts and their
Even those who are more inclined towards fundamental side can also benefit from Dow theory, as it can hint start & end of Bull/Bear runs thereby indication entry and exits.
Next basic is Wyckoff's Theory. It tells how accumulation and distribution happens with regularity and how the market actually
Dow theory is old but
Here I will share what I believe are essentials for anybody who is interested in stock markets and the resources to learn them, its from my experience and by no means exhaustive..
First the very basic : The Dow theory, Everybody must have basic understanding of it and must learn to observe High Highs, Higher Lows, Lower Highs and Lowers lows on charts and their
Even those who are more inclined towards fundamental side can also benefit from Dow theory, as it can hint start & end of Bull/Bear runs thereby indication entry and exits.

Next basic is Wyckoff's Theory. It tells how accumulation and distribution happens with regularity and how the market actually
Dow theory is old but
Old is Gold....
— Professor (@DillikiBiili) January 23, 2020
this Bharti Airtel chart is a true copy of the Wyckoff Pattern propounded in 1931....... pic.twitter.com/tQ1PNebq7d