Milad Rahimi

You also mentioned this paper: "Keep your Eyes on the Lane: Real-time Attention-guided Lane Detection" two times. In 2021 and 2020 sections. but with a subtle different title.

The link for this one: "PolyLaneNet: Lane Estimation via Deep Polynomial Regression" Should be: https://arxiv.org/abs/2004.10924

Like in the Faster R-CNN(that uses NMS during the training process), it is used to reduce the number of proposals and as a result, reduces computations.

I guessed the problem is related to the inverse function of the vectors, so I run this test to check the validity of inverse functions: ``` from torchhd import functional...

Have a look at this figure:  It shows the relationship between the similarity of two vectors and how much their division(x bind inv(y)) is close to identity. When this...

I was reading "[Learning with Holographic Reduced Representations](https://arxiv.org/pdf/2109.02157)", the paper introduced a better version of inverse function for HRR. I replaced the inverse function of torchhd with their optimized version...

Thank you, Mr. @rgayler for your insightful response. > I also believe that HRR and FHRR are equivalent (because you can losslessly transform between them) This clue was great. In...

The [level-hypervectors](https://arxiv.org/pdf/2205.07920) have an interesting feature: if we assume L'[i] like: > L'[i] = L[1] * L[-1] * inv(L[i]) Note: L'[1] is a vector that the second half of the...