FU Jun

@PatrykChrabaszcz Hello, after the solution is found quickly, the new weights will all be based on that solution.

@PatrykChrabaszcz No, whenever it reachs the goal state, the influence will be big and immediate to the following biased and random weights. Because it's reward is huge compared with other...

It's not average. It's only based on (R_positive_rank-R_negative_rank)/number_of_rewards, so the huge reward = 1, the tiny reward is 0.0000001. They are independent. On 11 May 2017 at 19:22, Patryk Chrabaszcz...

@sahiliitm Hello, in the paper, it is stated " We used the same preprocessing and feedforward CNN architecture used by (Mnih et al., 2016)". So it should be the traditional...

Me too I also find that skip connections don't help. It's unresponsible for the papers talk about it. Just a waste of our time.

I got in the same trouble, with pixels not filled. But sometimes, with luck, a new trained model can fix it. Please help.  @fregu856

@Girisha10 hi my friend have you found the reason and get the solution please?

Hi, @Ashwinning, Could you try this way: At Initializer::checkRT : 890, put the nGood inside the if{}: if(cosParallax

@dgsmith Maybe you can find something on https://github.com/raulmur/ORB_SLAM2/issues/66 . Did you try to initialize by homography (which means by 2d planar such as walls or grounds?), there is a bug...

Thank you for reply. And why an embedding for encoding, but no embedding for decoding? On 6 February 2017 at 12:00, kilickaya wrote: > Yes you can use. You can...