This extended abstract describes our solution for the Traffic4Cast Challenge 2019. The task requires modeling both fine-grained (pixel-level) and coarse (region-level) spatial structure while preserving temporal relationships across long sequences. Building on Conv-LSTM ideas, we introduce a tile-aware, cascaded-memory Conv-LSTM augmented with cross-frame additive attention and a memory-flexible training scheme: frames are sampled per spatial tile so the model learns tile-local dynamics and per-tile memory cells can be updated sparsely, paged, or compressed to scale to large maps. We provide a compact theoretical analysis (tight softmax/attention Lipschitz bound and a tiling error lower bound) explaining stability and the memory-accuracy tradeoffs, and empirically demonstrate improved scalability and competitive forecasting performance on large-scale traffic heatmaps.

Vision transformers have established a precedent of patchifying images into uniformly-sized chunks before processing. We hypothesize that this design choice may limit models in learning comprehensive and compositional representations from visual data. This paper explores the notion of providing semantically-meaningful visual tokens to transformer encoders within a vision-language pre-training framework. Leveraging off-the-shelf segmentation and scene-graph models, we extract representations of instance segmentation masks (referred to as tangible tokens) and relationships and actions (referred to as intangible tokens). Subsequently, we pre-train a vision-side transformer by incorporating these newly extracted tokens and aligning the resultant embeddings with caption embeddings from a text-side encoder. To capture the structural and semantic relationships among visual tokens, we introduce additive attention weights, which are used to compute self-attention scores. Our experiments on COCO demonstrate notable improvements over ViTs in learned representation quality across text-to-image (+47%) and image-to-text retrieval (+44%) tasks. Furthermore, we showcase the advantages on compositionality benchmarks such as ARO (+18%) and Winoground (+10%).

To respond to the challenge of efficient, automated violence detection from video, effective computer vision In this paper we introduce CUE-Net, a novel architecture methods are required. Deep learning techniques such designed for automated violence detection in video as Convolutional Neural Networks (CNNs) and more recently surveillance. As surveillance systems become more prevalent Transformer-based architectures have shown a great due to technological advances and decreasing costs, promise in solving computer vision related automated violence the challenge of efficiently monitoring vast amounts of detection [21, 22, 31]. The success of violence detection video data has intensified. CUE-Net addresses this challenge is highly dependent on the objects and people present by combining spatial Cropping with an enhanced version in the captured videos [22, 31]. Detection is difficult when of the UniformerV2 architecture, integrating convolutional the relevant features of the violent incidents are not captured and self-attention mechanisms alongside a novel properly, for example when the people involved in the Modified Efficient Additive Attention mechanism (which reduces violent incident are far away and occupy only a small part the quadratic time complexity of self-attention) to of the frame, as seen in one of the example videos from effectively and efficiently identify violent activities.

Object detection is a core component of perception systems, providing the ego vehicle with information about its surroundings to ensure safe route planning. While cameras and Lidar have significantly advanced perception systems, their performance can be limited in adverse weather conditions. In contrast, millimeter-wave technology enables radars to function effectively in such conditions. However, relying solely on radar for building a perception system doesn't fully capture the environment due to the data's sparse nature. To address this, sensor fusion strategies have been introduced. We propose a dual-branch framework to integrate radar and Lidar data for enhanced object detection. The primary branch focuses on extracting radar features, while the auxiliary branch extracts Lidar features. These are then combined using additive attention. Subsequently, the integrated features are processed through a novel Parallel Forked Structure (PFS) to manage scale variations. A region proposal head is then utilized for object detection. We evaluated the effectiveness of our proposed method on the Radiate dataset using COCO metrics. The results show that it surpasses state-of-the-art methods by $1.89\%$ and $2.61\%$ in favorable and adverse weather conditions, respectively. This underscores the value of radar-Lidar fusion in achieving precise object detection and localization, especially in challenging weather conditions.

Transformers have become the dominant model class in the last few years for large data, but their quadratic complexity in terms of sequence length has plagued them until now. Fastformer claims to be the fastest and most performant linear attention variant, able to consume long contexts at once. This is achieved by a combination of additive attention and elementwise products. OUTLINE: 0:00 - Intro & Outline 2:15 - Fastformer description 5:20 - Baseline: Classic Attention 10:00 - Fastformer architecture 12:50 - Additive Attention 18:05 - Query-Key element-wise multiplication 21:35 - Redundant modules in Fastformer 25:00 - Problems with the architecture 27:30 - Is this even attention? Abstract: Transformer is a powerful model for text understanding.

Based on an exponentially increasing number of academic articles, discovering and citing comprehensive and appropriate resources has become a non-trivial task. Conventional citation recommender methods suffer from severe information loss. For example, they do not consider the section of the paper that the user is writing and for which they need to find a citation, the relatedness between the words in the local context (the text span that describes a citation), or the importance on each word from the local context. These shortcomings make such methods insufficient for recommending adequate citations to academic manuscripts. In this study, we propose a novel embedding-based neural network called "dual attention model for citation recommendation (DACR)" to recommend citations during manuscript preparation. Our method adapts embedding of three dimensions of semantic information: words in the local context, structural contexts, and the section on which a user is working. A neural network is designed to maximize the similarity between the embedding of the three input (local context words, section and structural contexts) and the target citation appearing in the context. The core of the neural network is composed of self-attention and additive attention, where the former aims to capture the relatedness between the contextual words and structural context, and the latter aims to learn the importance of them. The experiments on real-world datasets demonstrate the effectiveness of the proposed approach.