Recent advances in visual activity recognition have raised the possibility of applications such as automated video surveillance. Effective approaches for such problems however require the ability to recognize the plans of agents from video information. Although traditional plan recognition algorithms depend on access to sophisticated planning domain models, one recent promising direction involves learning approximated (or shallow) domain models directly from the observed activity sequences DUP. One limitation is that such approaches expect observed action sequences as inputs. In many cases involving vision/sensing from raw data, there is considerable uncertainty about the specific action at any given time point. The most we can expect in such cases is probabilistic information about the action at that point. The input will then be sequences of such observed action distributions. In this work, we address the problem of constructing an effective data-interface that allows a plan recognition module to directly handle such observation distributions. Such an interface works like a bridge between the low-level perception module, and the high-level plan recognition module. We propose two approaches. The first involves resampling the distribution sequences to single action sequences, from which we could learn an action affinity model based on learned action (word) embeddings for plan recognition. The second is to directly learn action distribution embeddings by our proposed Distr2vec (distribution to vector) model, to construct an affinity model for plan recognition.
Decentralized reasoning is receiving increasing attention due to the distributed nature of knowledge on the Web. We address the problem of answering queries to distributed propositional reasoners which may be mutually inconsistent. This paper provides a formal characterization of a prioritized peerto-peer query answering framework that exploits a priority ordering over the peers, as well as a distributed entailment relation as an extension to established work on argumentation frameworks. We develop decentralized algorithms for computing query answers according to distributed entailment and prove their soundness and completeness. To improve the efficiency of query answering, we propose an ordering heuristic that exploits the peers' priority ordering and empirically evaluate its effectiveness.
Andrea James may not have a name as instantly recognizable as Laverne Cox or Caitlyn Jenner, but there are few activists who have done as much to connect, educate and help the trans community as this writer, producer and educator has over the past several decades. James, along with other prominent activists like Kate Bornstein, used the internet to connect trans and gender-nonconforming people to facilitate discussions about trans experiences long before the online space became regulated or corporatized. Her own site, Transsexual Road Map, became one of the first websites to focus on and educate individuals on the practical aspects of transition. Calling herself a consumer activist who works with trans issues, much of James' work has been out of the public eye but monumental in its significance for the queer community. In this interview with The Huffington Post, James reflects on her journey from rural Indiana to where she is today, the current state of trans politics and her legacy as a writer, producer, educator and so much more.
Whether it's Father's Day, Mother's Day, or some kind of International Marketing Research Firm branded holiday, nothing is more irritating than being forced to buy a gift for someone who insists they don't want a gift. Of course they want a gift. We all want shit we don't need. They, and especially our dads, just don't know how to ask for what they want. Most Father's Day gift guides are geared to please the Midwestern Cis Heterosexual Bear Dad: a man who likes meat, Android phones, sitting outside the changing room while his wife shops, overpriced gadgets from Brookstone, fishing, and "telling it like it is."