Conventional Bayesian Neural Networks (BNNs) are known to be capable of providing multiple outputs for a single input, the variations in which can be utilised to detect Out of Distribution (OOD) inputs. BNNs are difficult to train due to their sensitivity towards the choice of priors. To alleviate this issue, we propose pseudo-BNNs where instead of learning distributions over weights, we use point estimates and perturb weights at the time of inference. We modify the cost function of conventional BNNs and use it to learn parameters for the purpose of injecting right amount of random perturbations to each of the weights of a neural network with point estimate. In order to effectively segregate OOD inputs from In Distribution (ID) inputs using multiple outputs, we further propose two measures, derived from the index of dispersion and entropy of probability distributions, and combine them with the proposed pseudo-BNNs. Overall, this combination results in a principled technique to detect OOD samples at the time of inference. We evaluate our technique on a wide variety of neural network architectures and image classification datasets. We observe that our method achieves state of the art results and beats the related previous work on various metrics such as FPR at 95% TPR, AUROC, AUPR and Detection Error by just using 2 to 5 samples of weights per input.
翻译:众所周知,Bayesian Neural Networks(BNNs)能够为单一输入提供多种产出,这种变量可用于检测分配(OOD)输入的重量。BNNs由于对选择前期的敏感度而难以培训。为了缓解这一问题,我们提议假的BNS, 而不是学习超重分布, 我们在推断时使用点估计值和扰动权重; 我们修改常规的BNS的成本功能, 并用它来学习参数, 目的是为神经网络中带有点估测的每个重量注入适当数量的随机扰动参数。 为了有效地将OOOD输入从“ID”输入的多重输出中分离出来, 我们进一步提出两个衡量尺度, 取自分散指数和概率分布的酶, 并与拟议的伪-BNNS(BNS) 相结合。 总体而言,这种组合的结果是,在推断时采用一种有原则的技术,用于检测OD样本。 我们评估了在一系列的神经网络结构结构结构上采用的技术, 并用各种图解的精确度, 在先前的TR AS 5 中,我们用了一种方法, 在以往的图解中, 5 的模型中,我们用了一种方法, 的模型中, 实现了, 和图解的图解了我们用以前的图解的图解的图解 。