Temporal action detection (TAD) aims to identify and localize action instances in untrimmed videos, which is essential for various video understanding tasks. However, recent improvements in model performance, driven by larger feature extractors and datasets, have led to increased computational demands. This presents a challenge for applications like autonomous driving and robotics, which rely on limited computational resources. While existing channel pruning methods can compress these models, reducing the number of channels often hinders the parallelization efficiency of GPU, due to the inefficient multiplication between small matrices. Instead of pruning channels, we propose a Progressive Block Drop method that reduces model depth while retaining layer width. In this way, we still use large matrices for computation but reduce the number of multiplications. Our approach iteratively removes redundant blocks in two steps: first, we drop blocks with minimal impact on model performance; and second, we employ a parameter-efficient cross-depth alignment technique, fine-tuning the pruned model to restore model accuracy. Our method achieves a 25% reduction in computational overhead on two TAD benchmarks (THUMOS14 and ActivityNet-1.3) to achieve lossless compression. More critically, we empirically show that our method is orthogonal to channel pruning methods and can be combined with it to yield further efficiency gains.

Abstraction is a critical technique in general problem-solving, allowing complex tasks to be decomposed into smaller, manageable sub-tasks. While traditional symbolic planning relies on predefined primitive symbols to construct structured abstractions, its reliance on formal representations limits applicability to real-world tasks. On the other hand, reinforcement learning excels at learning end-to-end policies directly from sensory inputs in unstructured environments but struggles with compositional generalization in complex tasks with delayed rewards. In this paper, we propose Abductive Abstract Reinforcement Learning (A2RL), a novel neuro-symbolic RL framework bridging the two paradigms based on Abductive Learning (ABL), enabling RL agents to learn abstractions directly from raw sensory inputs without predefined symbols. A2RL induces a finite state machine to represent high-level, step-by-step procedures, where each abstract state corresponds to a sub-algebra of the original Markov Decision Process (MDP). This approach not only bridges the gap between symbolic abstraction and sub-symbolic learning but also provides a natural mechanism for the emergence of new symbols. Experiments show that A2RL can mitigate the delayed reward problem and improve the generalization capability compared to traditional end-to-end RL methods.

Temporal action detection (TAD) aims to locate action positions and recognize action categories in long-term untrimmed videos. Although many methods have achieved promising results, their robustness has not been thoroughly studied. In practice, we observe that temporal information in videos can be occasionally corrupted, such as missing or blurred frames. Interestingly, existing methods often incur a significant performance drop even if only one frame is affected. To formally evaluate the robustness, we establish two temporal corruption robustness benchmarks, namely THUMOS14-C and ActivityNet-v1.3-C. In this paper, we extensively analyze the robustness of seven leading TAD methods and obtain some interesting findings: 1) Existing methods are particularly vulnerable to temporal corruptions, and end-to-end methods are often more susceptible than those with a pretrained feature extractor; 2) Vulnera-bility mainly comes from localization error rather than classification error; 3) When corruptions occur in the middle of an action instance, TAD models tend to yield the largest performance drop. Besides building a benchmark, we further develop a simple but effective robust training method to defend against temporal corruptions, through the Frame-Drop augmentation and Temporal-Robust Consistency loss. Remarkably, our approach not only improves robustness but also yields promising improvements on clean data. We believe that this study will serve as a benchmark for future research in robust video analysis. Source code and models are available at https://github.com/Alvin-Zeng/temporal-robustness-benchmark.

Distance is a fundamental concept in spatial sciences. Spatial distance is a very important parameter to measure the relative positions between spatial objects and to indicate the degree of similarity between neighbouring objects. Indeed, spatial distance plays an important role in many areas such as neighbourhood analysis, structural similarity measure, image (or object) matching, clustering analysis, and so on. In this paper, existing computational models for the distance between spatial objects are evaluated and their problems pointed out; then, the concept of the Hausdorff distance is introduced as a metric indicator for different types of spatial objects. This distance is extended to a uniform representation by the introduction of the quantile, leading to the extended Hausdorff distance. Indeed, the so‐called extended Hausdorff distance is, in fact, a kind of metric characterized by the minimum distance, the Hausdorff distance, and the median Hausdorff distance. The first two can be used for measuring the dispersion and the last one for measuring the central tendency of the distance distribution between spatial objects. A method termed the ε‐buffer has been proposed for the computation of the median Hausdorff distance. Finally, potential applications are discussed.

This paper proposes a fast algorithm to enumerate all convex subgraphs that satisfy the I/O constraints from the dataflow graph (DFG) of a basic block. The algorithm can be tuned to determine all subgraphs or only those connected subgraphs. This allows a choice between better instruction-set extension (ISE) and faster design space exploration. The algorithm uses a grading method to identify the next node for inclusion into a subgraph. If the selected node is included, other related nodes are included as well, thus ensuring that the resultant subgraph is always convex and at the same time, reducing the problem size by a block of nodes. If the selected node is not included, the DFG will be split into smaller DFGs, thus reducing also the problem size. With this as base, the algorithm employs a simple but efficient method to prune the invalid subgraphs that violate the I/O constraints. Results show that for relatively small DFGs with small exploration space, the new algorithm has similar runtimes to that of existing algorithms. However, for larger DFGs with much larger exploration space and with multiple input and output constraints, the runtime improvement can be orders of magnitude better than that of existing algorithms. The new algorithm can be used to quickly identify custom instructions for ISE of embedded processors