We introduce the problem of random symmetric private information retrieval (RSPIR). In canonical PIR, a user downloads a message out of $K$ messages from $N$ non-colluding and replicated databases in such a way that no database can know which message the user has downloaded (user privacy). In SPIR, the privacy is symmetric, in that, not only that the databases cannot know which message the user has downloaded, the user itself cannot learn anything further than the particular message it has downloaded (database privacy). In RSPIR, different from SPIR, the user does not have an input to the databases, i.e., the user does not pick a specific message to download, instead is content with any one of the messages. In RSPIR, the databases need to send symbols to the user in such a way that the user is guaranteed to download a message correctly (random reliability), the databases do not know which message the user has received (user privacy), and the user does not learn anything further than the one message it has received (database privacy). This is the digital version of a blind box, also known as gachapon, which implements the above specified setting with physical objects for entertainment. This is also the blind version of $1$-out-of-$K$ oblivious transfer (OT), an important cryptographic primitive. We study the information-theoretic capacity of RSPIR for the case of $N=2$ databases. We determine its exact capacity for the cases of $K = 2, 3, 4$ messages. While we provide a general achievable scheme that is applicable to any number of messages, the capacity for $K\geq 5$ remains open.
Vision Transformers (ViTs) have gained significant popularity in recent years and have proliferated into many applications. However, it is not well explored how varied their behavior is under different learning paradigms. We compare ViTs trained through different methods of supervision, and show that they learn a diverse range of behaviors in terms of their attention, representations, and downstream performance. We also discover ViT behaviors that are consistent across supervision, including the emergence of Offset Local Attention Heads. These are self-attention heads that attend to a token adjacent to the current token with a fixed directional offset, a phenomenon that to the best of our knowledge has not been highlighted in any prior work. Our analysis shows that ViTs are highly flexible and learn to process local and global information in different orders depending on their training method. We find that contrastive self-supervised methods learn features that are competitive with explicitly supervised features, and they can even be superior for part-level tasks. We also find that the representations of reconstruction-based models show non-trivial similarity to contrastive self-supervised models. Finally, we show how the "best" layer for a given task varies by both supervision method and task, further demonstrating the differing order of information processing in ViTs.
Photo-identification (photo-id) is one of the main non-invasive capture-recapture methods utilised by marine researchers for monitoring cetacean (dolphin, whale, and porpoise) populations. This method has historically been performed manually resulting in high workload and cost due to the vast number of images collected. Recently automated aids have been developed to help speed-up photo-id, although they are often disjoint in their processing and do not utilise all available identifying information. Work presented in this paper aims to create a fully automatic photo-id aid capable of providing most likely matches based on all available information without the need for data pre-processing such as cropping. This is achieved through a pipeline of computer vision models and post-processing techniques aimed at detecting cetaceans in unedited field imagery before passing them downstream for individual level catalogue matching. The system is capable of handling previously uncatalogued individuals and flagging these for investigation thanks to catalogue similarity comparison. We evaluate the system against multiple real-life photo-id catalogues, achieving mAP@IOU[0.5] = 0.91, 0.96 for the task of dorsal fin detection on catalogues from Tanzania and the UK respectively and 83.1, 97.5% top-10 accuracy for the task of individual classification on catalogues from the UK and USA.
Witnessing the impressive achievements of pre-training techniques on large-scale data in the field of computer vision and natural language processing, we wonder whether this idea could be adapted in a grab-and-go spirit, and mitigate the sample inefficiency problem for visuomotor driving. Given the highly dynamic and variant nature of the input, the visuomotor driving task inherently lacks view and translation invariance, and the visual input contains massive irrelevant information for decision making, resulting in predominant pre-training approaches from general vision less suitable for the autonomous driving task. To this end, we propose PPGeo (Policy Pre-training via Geometric modeling), an intuitive and straightforward fully self-supervised framework curated for the policy pretraining in visuomotor driving. We aim at learning policy representations as a powerful abstraction by modeling 3D geometric scenes on large-scale unlabeled and uncalibrated YouTube driving videos. The proposed PPGeo is performed in two stages to support effective self-supervised training. In the first stage, the geometric modeling framework generates pose and depth predictions simultaneously, with two consecutive frames as input. In the second stage, the visual encoder learns driving policy representation by predicting the future ego-motion and optimizing with the photometric error based on current visual observation only. As such, the pre-trained visual encoder is equipped with rich driving policy related representations and thereby competent for multiple visuomotor driving tasks. Extensive experiments covering a wide span of challenging scenarios have demonstrated the superiority of our proposed approach, where improvements range from 2% to even over 100% with very limited data. Code and models will be available at https://github.com/OpenDriveLab/PPGeo.
On-orbit close proximity operations involve robotic spacecraft maneuvering and making decisions for a growing number of mission scenarios demanding autonomy, including on-orbit assembly, repair, and astronaut assistance. Of these scenarios, on-orbit assembly is an enabling technology that will allow large space structures to be built in-situ, using smaller building block modules. However, robotic on-orbit assembly involves a number of technical hurdles such as changing system models. For instance, grappled modules moved by a free-flying "assembler" robot can cause significant shifts in system inertial properties, which has cascading impacts on motion planning and control portions of the autonomy stack. Further, on-orbit assembly and other scenarios require collision-avoiding motion planning, particularly when operating in a "construction site" scenario of multiple assembler robots and structures. These complicating factors, relevant to many autonomous microgravity robotics use cases, are tackled in the ReSWARM flight experiments as a set of tests on the International Space Station using NASA's Astrobee robots. RElative Satellite sWarming and Robotic Maneuvering, or ReSWARM, demonstrates multiple key technologies for close proximity operations and on-orbit assembly: (1) global long-horizon planning, accomplished using offline and online sampling-based planner options that consider the system dynamics; (2) on-orbit reconfiguration model learning, using the recently-proposed RATTLE information-aware planning framework; and (3) robust control tools to provide low-level control robustness using current system knowledge. These approaches are detailed individually and in an "on-orbit assembly scenario" of multi-waypoint tracking on-orbit. Additionally, detail is provided discussing the practicalities of hardware implementation and unique aspects of working with Astrobee in microgravity.
By incorporating the graph structural information into Transformers, graph Transformers have exhibited promising performance for graph representation learning in recent years. Existing graph Transformers leverage specific strategies, such as Laplacian eigenvectors and shortest paths of the node pairs, to preserve the structural features of nodes and feed them into the vanilla Transformer to learn the representations of nodes. It is hard for such predefined rules to extract informative graph structural features for arbitrary graphs whose topology structure varies greatly, limiting the learning capacity of the models. To this end, we propose an adaptive graph Transformer, termed Multi-Neighborhood Attention based Graph Transformer (MNA-GT), which captures the graph structural information for each node from the multi-neighborhood attention mechanism adaptively. By defining the input to perform scaled-dot product as an attention kernel, MNA-GT constructs multiple attention kernels based on different hops of neighborhoods such that each attention kernel can capture specific graph structural information of the corresponding neighborhood for each node pair. In this way, MNA-GT can preserve the graph structural information efficiently by incorporating node representations learned by different attention kernels. MNA-GT further employs an attention layer to learn the importance of different attention kernels to enable the model to adaptively capture the graph structural information for different nodes. Extensive experiments are conducted on a variety of graph benchmarks, and the empirical results show that MNA-GT outperforms many strong baselines.
Video Super-Resolution (VSR) aims to recover sequences of high-resolution (HR) frames from low-resolution (LR) frames. Previous methods mainly utilize temporally adjacent frames to assist the reconstruction of target frames. However, in the real world, there is a lot of irrelevant information in adjacent frames of videos with fast scene switching, these VSR methods cannot adaptively distinguish and select useful information. In contrast, with a transformer structure suitable for temporal tasks, we devise a novel adaptive scenario video super-resolution method. Specifically, we use optical flow to label the patches in each video frame, only calculate the attention of patches with the same label. Then select the most relevant label among them to supplement the spatial-temporal information of the target frame. This design can directly make the supplementary information come from the same scene as much as possible. We further propose a cross-scale feature aggregation module to better handle the scale variation problem. Compared with other video super-resolution methods, our method not only achieves significant performance gains on single-scene videos but also has better robustness on cross-scene datasets.
DNA-Encoded Library (DEL) technology has enabled significant advances in hit identification by enabling efficient testing of combinatorially-generated molecular libraries. DEL screens measure protein binding affinity though sequencing reads of molecules tagged with unique DNA-barcodes that survive a series of selection experiments. Computational models have been deployed to learn the latent binding affinities that are correlated to the sequenced count data; however, this correlation is often obfuscated by various sources of noise introduced in its complicated data-generation process. In order to denoise DEL count data and screen for molecules with good binding affinity, computational models require the correct assumptions in their modeling structure to capture the correct signals underlying the data. Recent advances in DEL models have focused on probabilistic formulations of count data, but existing approaches have thus far been limited to only utilizing 2-D molecule-level representations. We introduce a new paradigm, DEL-Dock, that combines ligand-based descriptors with 3-D spatial information from docked protein-ligand complexes. 3-D spatial information allows our model to learn over the actual binding modality rather than using only structured-based information of the ligand. We show that our model is capable of effectively denoising DEL count data to predict molecule enrichment scores that are better correlated with experimental binding affinity measurements compared to prior works. Moreover, by learning over a collection of docked poses we demonstrate that our model, trained only on DEL data, implicitly learns to perform good docking pose selection without requiring external supervision from expensive-to-source protein crystal structures.
Domain shift across crowd data severely hinders crowd counting models to generalize to unseen scenarios. Although domain adaptive crowd counting approaches close this gap to a certain extent, they are still dependent on the target domain data to adapt (e.g. finetune) their models to the specific domain. In this paper, we aim to train a model based on a single source domain which can generalize well on any unseen domain. This falls into the realm of domain generalization that remains unexplored in crowd counting. We first introduce a dynamic sub-domain division scheme which divides the source domain into multiple sub-domains such that we can initiate a meta-learning framework for domain generalization. The sub-domain division is dynamically refined during the meta-learning. Next, in order to disentangle domain-invariant information from domain-specific information in image features, we design the domain-invariant and -specific crowd memory modules to re-encode image features. Two types of losses, i.e. feature reconstruction and orthogonal losses, are devised to enable this disentanglement. Extensive experiments on several standard crowd counting benchmarks i.e. SHA, SHB, QNRF, and NWPU, show the strong generalizability of our method.
Nowadays, the applications of hydraulic systems are present in a wide variety of devices in both industrial and everyday environments. The implementation and usage of hydraulic systems have been well documented; however, today, this still faces a challenge, the integration of tools that allow more accurate information about the functioning and operation of these systems for proactive decision-making. In industrial applications, many sensors and methods exist to measure and determine the status of process variables (e.g., flow, pressure, force). Nevertheless, little has been done to have systems that can provide users with device-health information related to hydraulic devices integrated into the machinery. Implementing artificial intelligence (AI) technologies and machine learning (ML) models in hydraulic system components has been identified as a solution to the challenge many industries currently face: optimizing processes and carrying them out more safely and efficiently. This paper presents a solution for the characterization and estimation of anomalies in one of the most versatile and used devices in hydraulic systems, cylinders. AI and ML models were implemented to determine the current operating status of these hydraulic components and whether they are working correctly or if a failure mode or abnormal condition is present.