Tweetovi

Blokirali ste korisnika/cu @sagarpath

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @sagarpath

  1. proslijedio/la je Tweet
    1. velj

    To the left, you see a trained agent playing a level of a game. To the right, you see the same playthrough from an agent-centric perspective: cropped, translated, and rotated with the agent in the center. Which perspective is the best input for the agent?

    Prikaži ovu nit
    Poništi
  2. proslijedio/la je Tweet
    31. sij

    BlenderProc generates visual data for training CNNs including depth, normals, semantic segmentation, pose annotations, and of course color images and much more! Code:

    Poništi
  3. proslijedio/la je Tweet
    31. sij

    The benefits of synthetic data for action categorization

    Poništi
  4. proslijedio/la je Tweet
    30. sij

    Humans learn from curriculum since birth. We can learn complicated math problems because we have accumulated enough prior knowledge. This could be true for training a ML/RL model as well. Let see how curriculum can help an RL agent learn:

    Poništi
  5. proslijedio/la je Tweet
    29. sij

    PolyGame: An open-source framework for training AI players through self-play. Deals with many games, board size variation, partial observability... Interesting generalization tidbit: It plays Go on 19x19 at very good level after training only on 13x13.

    Poništi
  6. proslijedio/la je Tweet
    29. sij

    ML Fairness Gym is a set of components for building simple simulations that explore the potential long-run impacts of deploying machine learning-based decision systems in social environments, based on OpenAI Gym interface.

    Poništi
  7. proslijedio/la je Tweet
    28. sij

    New paper: Towards a Human-like Open-Domain Chatbot. Key takeaways: 1. "Perplexity is all a chatbot needs" ;) 2. We're getting closer to a high-quality chatbot that can chat about anything Paper: Blog:

    Prikaži ovu nit
    Poništi
  8. proslijedio/la je Tweet
    28. sij

    In “Artificial Intelligence, Values and Alignment” DeepMind’s explores approaches to aligning AI with a wide range of human values:

    Poništi
  9. proslijedio/la je Tweet
    28. sij

    Check out Meena, a new state-of-the-art open-domain conversational agent, released along with a new evaluation metric, the Sensibleness and Specificity Average, which captures basic, but important attributes for normal conversation. Learn more below!

    Prikaži ovu nit
    Poništi
  10. proslijedio/la je Tweet
    28. sij

    What if we see level generation as a game, where each action changes the level in some way? Well, we could use RL to learn to "play" level generation! We introduce Procedural Content Generation via Reinforcement Learning (PCGRL), a new paradigm for PCG.

    Prikaži ovu nit
    Poništi
  11. proslijedio/la je Tweet
    28. sij

    Procedural Content Generation via Reinforcement Learning “A new approach to procedural content generation in games, where level design is framed as a game (as a sequential task problem), and the content generator itself is learned.”

    Poništi
  12. proslijedio/la je Tweet
    26. sij

    Quaternions and Euler angles are discontinuous and difficult for neural networks to learn. They show 3D rotations have continuous representations in 5D and 6D, which are more suitable for learning. i.e. regress two vectors and apply Graham-Schmidt (GS).

    Prikaži ovu nit
    Poništi
  13. proslijedio/la je Tweet
    24. sij

    Impressive results on single image depth estimation by Ranftl et al

    Prikaži ovu nit
    Poništi
  14. proslijedio/la je Tweet
    23. sij

    Q-learning is difficult to apply when the number of available actions is large. We show that a simple extension based on amortized stochastic search allows Q-learning to scale to high-dimensional discrete, continuous or hybrid action spaces:

    Poništi
  15. proslijedio/la je Tweet
    22. sij

    code showing the generic learning setup and reproducing simple experiments is now available! Code: Project Page: Paper:

    Poništi
  16. proslijedio/la je Tweet
    29. tra 2019.

    Excited to release our work on MetaSim which learns to generate synthetic datasets that resemble real datasets in content. Antonio Torralba etc paper: project page:

    Prikaži ovu nit
    Poništi
  17. proslijedio/la je Tweet
    22. sij

    Excited to share PCGrad, a super simple & effective method for multi-task learning & multi-task RL: project conflicting gradients On Meta-World MT50, PCGrad can solve *2x* more tasks than prior methods w/ Tianhe Yu, S Kumar, Gupta, ,

    Poništi
  18. proslijedio/la je Tweet
    21. sij

    We’re releasing a major update to Facebook AI’s open source AI Habitat platform for training embodied AI agents in photorealistic 3D virtual environments. AI Habitat now supports interactive objects, realistic physics modeling, and more.

    Poništi
  19. proslijedio/la je Tweet
    21. sij

    Facebook AI has effectively solved the task of point-goal navigation by AI agents in simulated environments, using only a camera, GPS, and compass data. Agents achieve 99.9% success in a variety of virtual settings, such as houses and offices.

    Poništi
  20. proslijedio/la je Tweet
    20. sij

    Given the smoothness of videos, can we learn models more efficiently than with ? We present Sideways - a step towards a high-throughput, approximate backprop that considers the one-way direction of time and pipelines forward and backward passes.

    Poništi

Čini se da učitavanje traje već neko vrijeme.

Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.

    Možda bi vam se svidjelo i ovo:

    ·