… We consider the classical problem of estimating the covariance matrix of a sub-Gaussian
distribution from iid samples in the novel context of coarse quantization, that is, instead of
having full knowledge of the samples, they are quantized to one or two bits per entry. This
problem occurs naturally in signal processing applications. We introduce new estimators in two
different … In our second quantization model, we aim to estimate the full covariance matrix of
any centered sub-Gaussian distribution. To achieve this, we introduce dithering in the one-bit …

We consider the classical problem of estimating the covariance matrix of a subgaussian
distribution from iid samples in the novel context of coarse quantization, ie, instead of having
full knowledge of the samples, they are quantized to one or two bits per entry. This problem
occurs naturally in signal processing applications. We introduce new estimators in two
different quantization scenarios and derive non‐asymptotic estimation error bounds in terms
of the operator norm. In the first scenario we consider a simple, scale‐invariant one‐bit …