Not mine. A mother walking with her daughter in Bradford is perfectly justified in changing course when she sees a group of Pakistani Muslims happening upon her path.
I don't actually consider it racism to treat race as a data point that can add information to a Bayesian analysis of a situation. There are statistical trends, and knowing those trends can actually tell you
something. As an individual I don't have a problem with using that information.
There are a couple of issues though. One is that other information can be much more robust and informative, and should supersede whatever initial assumptions were made. If you don't update your priors based on new information you're doing a bad job of anylising the situation.
Another issue is that if institutions use race as a data point they will tend to perpetuate the situation that leads to those differences. I know you don't actually agree with this because we disagree about the causes here.
A third issue is that some of those statistics may actually not reflect the reality of the situation. I'm skeptical that all of the differences in crime rates are down to judicial bias, but I wouldn't be surprised if there's some effect. That's a difficult effect to estimate, so how much should you alter your estimation of the actual rate based on it? Tough question.
Another issue is that it's possible that reactions to those perceived situations are actually a causal factor. By crossing the street when you see a group of young Pakistani men, you show fear and make yourself a target. The same would have happened with groups of young men of other races, but people don't actually cross the street when they see them, so it doesn't happen as often, and as such the crime rate goes up. The perception leads to the reality. (I'm not saying this mechanism is actually at play, but it is possible for stereotypes to perpetuate themselves and it's going to be hard to tease them out in the data).
For all of these reasons I think that in most situations we should all take much less notice of race than we actually do. The situation that you mention may be an exception, but for some of the reasons mentioned above I do think police should be color-blind.