http://algs4.cs.princeton.edu/home/ is my favorite. The implementations of the algorithms are really nice (quite a feat, considering the lang is Java :-p) and each section builds on top of a previous one, when needed (e.g. many algorithms require a Symbol Table, which is built in one of the chapters of the first volume). The resulting implementations are incredible terse but at the same time clean and easy to understand!

Cormen's book is a bit more thedious to read and I don't know it that well but I think is a little bit more rigorous with math. Sedgewick's book doesn't mention asymptotic limits or BigO notation, rather talks about the run time / memory requirements in more informal terms, like "run time being proportional to the square of the number of input items", etc. But, there's another book from Sedgewick (and Flajolet!) if you want to go deeper into analysis: (1).

Skienna's book is great to have a wider outline on all sorts of algorithms, including some less widely known and (I think) not covered in either of Sedgewick's or Cormen's book. It is, if I'm allowed the pun, more of a breadth first than a depth first approach to the subject.

1: http://www.amazon.com/Introduction-Analysis-Algorithms-2nd/d...