![\"\"](\"https:\/\/mscvprojects.ri.cmu.edu\/2022team1\/wp-content\/uploads\/sites\/56\/2022\/04\/Stroller-1-1024x427.png\")
However, the LPS setup fails to consider realistic testing environments. First, in the real-world, there may be constant distribution shifts and LPS does not account for this. Second, and more crucially, the network is only trained to segment regions which belong to a predefined vocabulary of K<\/em> classes. However, if a rare object such as a stroller is encountered (Fig 1), the network is not trained to recognize this, which may lead to catastrophic consequences. In an ideal situation, the network should alert such points as unknown<\/em> or other<\/em> (i.e., not seen in training set). <\/p>\n\n\n\n The ability to segment these unseen classes is very important for safety-critical applications such as autonomous navigation. Additionally, it may also be useful for data-labelling, especially in the long-tail.<\/p>\n\n\n\n To overcome the challenges of using LPS methods in a real-world setting, we propose a new task which extends LPS in an open-world (LiPSOW). Under LiPSOW, methods are evaluated on a test-set which have a different data distribution than the training set. Therefore, to do well on LiPSOW, the method should be capable of handling shifting domains. Moreover, to account for novel objects from the long-tail, we evaluate algorithms by their ability to recognize other<\/em> points and segment instances from other<\/em>. In summary, in addition to stuff<\/em> and thing<\/em> classes (as is the case in LPS), we introduce an other<\/em> class, which may internally consist of stuff<\/em> or things<\/em>, and LiPSOW methods must segment instances from both thing<\/em> and other<\/em> classes.<\/p>\n\n\n\n In this work, our contributions are three-fold:<\/p>\n\n\n\nProblem Statement<\/h2>\n\n\n\n
![\"\"](\"https:\/\/mscvprojects.ri.cmu.edu\/2022team1\/wp-content\/uploads\/sites\/56\/2022\/12\/Screenshot-2022-12-15-at-11.47.44-PM-1024x247.png\")
Right:<\/strong> In contrast, LiPSOW requires novel objects to be segmented as other<\/em><\/figcaption><\/figure>\n\n\n\nOur Contributions<\/h2>\n\n\n\n