Automated recognition of facial expressions is an important problem in computer vision applications. Due to the vagueness in class definitions, expression recognition is often conceived as a fuzzy label problem. Annotating a data point in such a problem involves significant manual effort. Active learning techniques are effective in reducing human labeling effort to induce a classification model as they automatically select the salient and exemplar instances from vast amounts of unlabeled data. Further, to address the high redundancy in data such as image or video sequences as well as to account for the presence of multiple labeling agents, there have been recent attempts towards a batch mode form of active learning where a batch of data points is selected simultaneously from an unlabeled set. In this paper, we propose a novel optimization-based batch mode active learning technique for fuzzy label classification problems. To the best of our knowledge, this is the first effort to develop such a scheme primarily intended for the fuzzy label context. The proposed algorithm is computationally simple, easy to implement and has provable performance bounds. Our results on facial expression datasets corroborate the efficacy of the framework in reducing human annotation effort in real world recognition applications involving fuzzy labels.