Retrieving images with objects that are semantically similar to objects of interest (OOI) in a query image has many practical use cases. A few examples include fixing failures like false negatives/positives of a learned model or mitigating class imbalance in a dataset. The targeted selection task requires finding the relevant data from a large-scale pool of unlabeled data. Manual mining at this scale is infeasible. Further, the OOI are often small and occupy less than 1% of image area, are occluded, and co-exist with many semantically different objects in cluttered scenes. Existing semantic image retrieval methods often focus on mining for larger sized geographical landmarks, and/or require extra labeled data, such as images/image-pairs with similar objects, for mining images with generic objects. We propose a fast and robust template matching algorithm in the DNN feature space, that retrieves semantically similar images at the object-level from a large unlabeled pool of data. We project the region(s) around the OOI in the query image to the DNN feature space for use as the template. This enables our method to focus on the semantics of the OOI without requiring extra labeled data. In the context of autonomous driving, we evaluate our system for targeted selection by using failure cases of object detectors as OOI. We demonstrate its efficacy on a large unlabeled dataset with 2.2M images and show high recall in mining for images with small-sized OOI. We compare our method against a well-known semantic image retrieval method, which also does not require extra labeled data. Lastly, we show that our method is flexible and retrieves images with one or more semantically different co-occurring OOI seamlessly.

Computational biology and bioinformatics provide vast data gold-mines from protein sequences, ideal for Language Models (LMs) taken from Natural Language Processing (NLP). These LMs reach for new prediction frontiers at low inference costs. Here, we trained two auto-regressive models (Transformer-XL, XLNet) and four auto-encoder models (BERT, Albert, Electra, T5) on data from UniRef and BFD containing up to 393 billion amino acids. The protein LMs (pLMs) were trained on the Summit supercomputer using 5616 GPUs and TPU Pod up-to 1024 cores. Dimensionality reduction revealed that the raw pLM-embeddings from unlabeled data captured some biophysical features of protein sequences. We validated the advantage of using the embeddings as exclusive input for several subsequent tasks: (1) a per-residue (per-token) prediction of protein secondary structure (3-state accuracy Q3=81%-87%); (2) per-protein (pooling) predictions of protein sub-cellular location (ten-state accuracy: Q10=81%) and membrane versus water-soluble (2-state accuracy Q2=91%). For secondary structure, the most informative embeddings (ProtT5) for the first time outperformed the state-of-the-art without multiple sequence alignments (MSAs) or evolutionary information thereby bypassing expensive database searches. Taken together, the results implied that pLMs learned some of the grammar of the language of life. All our models are available through https://github.com/agemagician/ProtTrans.

An open problem in slotted Aloha protocols is to optimally estimate the number of participants as such knowledge is crucial to select the optimal frame length. First results are known in literature based on observing the slot fill levels in case of empty slots and single occupancies (singleton slots). Advances in signal processing allow now also to decode successfully slots with higher fill levels, for example, due to multiple antennas. In this paper we derive the maximum likelihood estimator when arbitrary occupancies up to a maximal fill level R have been observed. Due to our novel approach, the derivation is rather simple and its implementation is of low complexity.

Java was designed as a secure language that supports running untrusted code as part of trusted applications. For safety reasons, Java therefore defines a memory model that prevents undefined behavior in multi-threaded programs even if the programs are not correctly synchronized. Because of the potential negative performance impact the Java designers did not choose a simple and natural memory model, such as sequential consistency, but instead developed a relaxed memory model that gives the compiler more optimization opportunities. As it is today, however, the relaxed Java Memory Model is not only hard to understand but it unnecessarily complicates reasoning about parallel programs and it turned out to be difficult to implement correctly. This paper presents an optimizing compiler for a Java version that has sequential consistency as its memory model. Based on a programming model with explicit happens-before constraints between tasks, we describe a static schedule analysis that computes whether two tasks may be executed in parallel or if they are ordered. During optimization, the task-ordering information is exploited to reduce the number of volatile memory accesses the compiler must insert to guarantee sequential consistency. The evaluation shows that scheduling information significantly improves the effectiveness of the optimizations. For our set of multi-threaded benchmarks the fully optimizing compiler removes between 70% and 100% of the volatile memory accesses inserted by the non-optimizing compiler. As a result, the overhead of sequentially consistent Java compared to standard Java is reduced from 136% on average for the unoptimized version to 11% on average for the optimized version. The results indicate that with appropriate optimizations, sequential consistency can be a feasible alternative to the Java Memory Model.

Arbitration and scheduling of multiple tags in state-of-the-art Radio Frequency Identification (RFID) systems is accomplished on the medium access control layer. Currently, only answers of a single tag can be decoded in such a system. If multiple tags respond simultaneously, a collision occurs. In that case, conventional systems discard the physical layer information and a retransmission is executed. This work shows how to recover from such collisions on the physical layer and successfully read the data. The contributions of the paper are: 1) An analysis of the achievable throughput increase of a system, that can recover from collisions at a physical layer is given. 2) A model for a description of collisions on the physical layer is presented. 3) Based on this model, we propose a channel estimation method and two types of receiver structures for separating the signals of a collision of two tags: first, single antenna receivers that discriminate the sources of the two tags in the I/Q plane, and second, multiple antenna receivers which exploit the different spatial signatures of both tags. 4) The functionality of the proposed receiver structures is verified with measurement data of two colliding tags. Eventually, a performance analysis of the receivers is provided.

Framed Slotted Aloha (FSA) is a popular anticollision technique in state-of-the-art RF-ID systems, as in ISO/IEC CD 18000-6 for 900MHz or the EPCglobal HF Gen 2 draft for 13.56MHz. In many applications the number of tags entering and leaving the detection range of the reader is subject to a strong fluctuation and usually unknown. The current number of tags in the field is a crucial parameter to operate the FSA anti-collision in an optimal manner. Therefore, a lot of effort is spent on the estimation of this parameter and a range of different estimation techniques exist. The contributions of this paper are: 1) a closed formula for the probability of any observed event defined by the number of empty, singleton, and collision slots in the observed frame is developed and empirically verified. 2) This formula is then modified to compute the probability for partly observed frames as well which is of great interest as the referred standards allow for the in-frame adjustment of the frame size without quitting the interrogation round. 3) Then, a maximum likelihood estimator is formulated to yield the estimated number of tags on a slot-wise basis. 4) Its superior estimation performance is compared to the known best estimators over the complete parameter set. While its performance is strongly superior compared to Schoute¿s estimate, compared to Vogt¿s MSE estimator only marginally improvement is obtained1.

Policy-based management (PBM) is a promising approach to realize adaptivity in networks that would be hard - if not impossible - for administrators to configure manually. In PBM systems, the network configuration is not directly expressed in terms of low-level properties, such as IP or MAC addresses; rather, complex business entities, such as users or departments, can be defined and allow the administrators to specify high-level network policies. Even though PBM can reduce the effort for managing complex networks significantly, today's solutions lack a generic way for expressing adaptive policies. A PBM system can be called adaptive if the decision of when and in which order policies are activated is not only based on fixed predefined dates but the system also takes environmental and internal events into account. In this paper we describe a model of adaptive timing constraints and introduce an architecture for executing timing specifications (sets of constraints). Within a timing specification, the administrator usually creates single constraints only implicitly by defining the lifetime of managed resources, sequences of events, and high- level temporal relationships between them. Because each timing constraint defines a < relation between two single events, a timing specification can be described as a partial ordering of events over the event space. During execution, a timing service triggers PBM systems to activate, inactivate, or adapt management policies according to the timing specification.