Video generative models pre-trained on large-scale internet datasets have achieved remarkable success, excelling at producing realistic synthetic videos. However, they often generate clips based on static prompts (e.g., text or images), limiting their ability to model interactive and dynamic scenarios. In this paper, we propose Dynamic World Simulation (DWS), a novel approach to transform pre-trained video generative models into controllable world simulators capable of executing specified action trajectories. To achieve precise alignment between conditioned actions and generated visual changes, we introduce a lightweight, universal action-conditioned module that seamlessly integrates into any existing model. Instead of focusing on complex visual details, we demonstrate that consistent dynamic transition modeling is the key to building powerful world simulators. Building upon this insight, we further introduce a motion-reinforced loss that enhances action controllability by compelling the model to capture dynamic changes more effectively. Experiments demonstrate that DWS can be versatilely applied to both diffusion and autoregressive transformer models, achieving significant improvements in generating action-controllable, dynamically consistent videos across games and robotics domains. Moreover, to facilitate the applications of the learned world simulator in downstream tasks such as model-based reinforcement learning, we propose prioritized imagination to improve sample efficiency, demonstrating competitive performance compared with state-of-the-art methods.

Distributed Deep Reinforcement Learning (DRL) aims to leverage more computational resources to train autonomous agents with less training time. Despite recent progress in the field, reproducibility issues have not been sufficiently explored. This paper first shows that the typical actor-learner framework can have reproducibility issues even if hyperparameters are controlled. We then introduce Cleanba, a new open-source platform for distributed DRL that proposes a highly reproducible architecture. Cleanba implements highly optimized distributed variants of PPO and IMPALA. Our Atari experiments show that these variants can obtain equivalent or higher scores than strong IMPALA baselines in moolib and torchbeast and PPO baseline in CleanRL. However, Cleanba variants present 1) shorter training time and 2) more reproducible learning curves in different hardware settings.

Deep reinforcement learning (RL) algorithms suffer severe performance degradation when the interaction data is scarce, which limits their real-world application. Recently, visual representation learning has been shown to be effective and promising for boosting sample efficiency in RL. These methods usually rely on contrastive learning and data augmentation to train a transition model, which is different from how the model is used in RL---performing value-based planning. Accordingly, the learned representation by these visual methods may be good for recognition but not optimal for estimating state value and solving the decision problem. To address this issue, we propose a novel method, called value-consistent representation learning (VCR), to learn representations that are directly related to decision-making. More specifically, VCR trains a model to predict the future state (also referred to as the "imagined state'') based on the current one and a sequence of actions. Instead of aligning this imagined state with a real state returned by the environment, VCR applies a Q value head on both of the states and obtains two distributions of action values. Then a distance is computed and minimized to force the imagined state to produce a similar action value prediction as that by the real state. We develop two implementations of the above idea for the discrete and continuous action spaces respectively. We conduct experiments on Atari 100k and DeepMind Control Suite benchmarks to validate their effectiveness for improving sample efficiency. It has been demonstrated that our methods achieve new state-of-the-art performance for search-free RL algorithms.

Existing imitation learning (IL) methods such as inverse reinforcement learning (IRL) usually have a double-loop training process, alternating between learning a reward function and a policy and tend to suffer long training time and high variance. In this work, we identify the benefits of differentiable physics simulators and propose a new IL method, i.e., Imitation Learning as State Matching via Differentiable Physics (ILD), which gets rid of the double-loop design and achieves significant improvements in final performance, convergence speed, and stability. The proposed ILD incorporates the differentiable physics simulator as a physics prior into its computational graph for policy learning. ILD unrolls the dynamics by sampling actions from a parameterized policy and minimizing the distance between the expert trajectory and the agent trajectory. It back-propagates the gradient into the policy via temporal physics operators, which improves the transferability to unseen environments and yields higher final performance. ILD has a single-loop structure that stabilizes and speeds up training. It dynamically selects learning objectives for each state during optimization to simplify the complex optimization land-scape. Experiments show that ILD outperforms state-of-the-art methods in continuous control tasks with Brax, and can be applied to deformable object manipulation tasks, generalized to unseen configurations.11The link to the code: https://github.com/sail-sg/ILD

The major challenge of offline RL is the distribution shift that appears when out-of-distribution actions are queried, which makes the policy improvement direction biased by extrapolation errors. Most existing methods address this problem by penalizing the policy or value for deviating from the behavior policy during policy improvement or evaluation. In this work, we propose a novel MISA framework to approach offline RL from the perspective of Mutual Information between States and Actions in the dataset by directly constraining the policy improvement direction. MISA constructs lower bounds of mutual information parameterized by the policy and Q-values. We show that optimizing this lower bound is equivalent to maximizing the likelihood of a one-step improved policy on the offline dataset. Hence, we constrain the policy improvement direction to lie in the data manifold. The resulting algorithm simultaneously augments the policy evaluation and improvement by adding mutual information regularizations. MISA is a general framework that unifies conservative Q-learning (CQL) and behavior regularization methods (e.g., TD3+BC) as special cases. We introduce 3 different variants of MISA, and empirically demonstrate that tighter mutual information lower bound gives better offline RL performance. In addition, our extensive experiments show MISA significantly outperforms a wide range of baselines on various tasks of the D4RL benchmark, e.g., achieving 742.9 total points on gym-locomotion tasks. Our code is attached and will be released upon publication.

There has been significant progress in developing reinforcement learning (RL) training systems. Past works such as IMPALA, Apex, Seed RL, Sample Factory, and others, aim to improve the system's overall throughput. In this paper, we aim to address a common bottleneck in the RL training system, i.e., parallel environment execution, which is often the slowest part of the whole system but receives little attention. With a curated design for paralleling RL environments, we have improved the RL environment simulation speed across different hardware setups, ranging from a laptop and a modest workstation, to a high-end machine such as NVIDIA DGX-A100. On a high-end machine, EnvPool achieves one million frames per second for the environment execution on Atari environments and three million frames per second on MuJoCo environments. When running EnvPool on a laptop, the speed is 2.8x that of the Python subprocess. Moreover, great compatibility with existing RL training libraries has been demonstrated in the open-sourced community, including CleanRL, rl_games, DeepMind Acme, etc. Finally, EnvPool allows researchers to iterate their ideas at a much faster pace and has great potential to become the de facto RL environment execution engine. Example runs show that it only takes five minutes to train agents to play Atari Pong and MuJoCo Ant on a laptop. EnvPool is open-sourced at https://github.com/sail-sg/envpool.

Travel-time prediction constitutes a task of high importance in transportation networks, with web mapping services like Google Maps regularly serving vast quantities of travel time queries from users and enterprises alike. Further, such a task requires accounting for complex spatiotemporal interactions (modelling both the topological properties of the road network and anticipating events---such as rush hours---that may occur in the future). Hence, it is an ideal target for graph representation learning at scale. Here we present a graph neural network estimator for estimated time of arrival (ETA) which we have deployed in production at Google Maps. While our main architecture consists of standard GNN building blocks, we further detail the usage of training schedule methods such as MetaGradients in order to make our model robust and production-ready. We also provide prescriptive studies: ablating on various architectural decisions and training regimes, and qualitative analyses on real-world situations where our model provides a competitive edge. Our GNN proved powerful when deployed, significantly reducing negative ETA outcomes in several regions compared to the previous production baseline (40+% in cities like Sydney).

Off-policy learning allows us to learn about possible policies of behavior from experience generated by a different behavior policy. Temporal difference (TD) learning algorithms can become unstable when combined with function approximation and off-policy sampling—this is known as the “deadly triad”. Emphatic temporal difference (ETD($\lambda$)) algorithm ensures convergence in the linear case by appropriately weighting the TD($\lambda$) updates. In this paper, we extend the use of emphatic methods to deep reinforcement learning agents. We show that naively adapting ETD($\lambda$) to popular deep reinforcement learning algorithms, which use forward view multi-step returns, results in poor performance. We then derive new emphatic algorithms for use in the context of such algorithms, and we demonstrate that they provide noticeable benefits in small problems designed to highlight the instability of TD methods. Finally, we observed improved performance when applying these algorithms at scale on classic Atari games from the Arcade Learning Environment.

Temporal abstractions in the form of options have been shown to help reinforcement learning (RL) agents learn faster. However, despite prior work on this topic, the problem of discovering options through interaction with an environment remains a challenge. In this paper, we introduce a novel meta-gradient approach for discovering useful options in multi-task RL environments. Our approach is based on a manager-worker decomposition of the RL agent, in which a manager maximises rewards from the environment by learning a task-dependent policy over both a set of task-independent discovered-options and primitive actions. The option-reward and termination functions that define a subgoal for each option are parameterised as neural networks and trained via meta-gradients to maximise their usefulness. Empirical analysis on gridworld and DeepMind Lab tasks show that: (1) our approach can discover meaningful and diverse temporally-extended options in multi-task RL domains, (2) the discovered options are frequently used by the agent while learning to solve the training tasks, and (3) that the discovered options help a randomly initialised manager learn faster in completely new tasks.

The objective of a reinforcement learning agent is to behave so as to maximise the sum of a suitable scalar function of state: the reward. These rewards are typically given and immutable. In this paper, we instead consider the proposition that the reward function itself can be a good locus of learned knowledge. To investigate this, we propose a scalable meta-gradient framework for learning useful intrinsic reward functions across multiple lifetimes of experience. Through several proof-of-concept experiments, we show that it is feasible to learn and capture knowledge about long-term exploration and exploitation into a reward function. Furthermore, we show that unlike policy transfer methods that capture “how” the agent should behave, the learned reward functions can generalise to other kinds of agents and to changes in the dynamics of the environment by capturing “what” the agent should strive to do.

Reinforcement learning (RL) algorithms update an agent’s parameters according to one of several possible rules, discovered manually through years of research. Automating the discovery of update rules from data could lead to more efficient algorithms, or algorithms that are better adapted to specific environments. Although there have been prior attempts at addressing this significant scientific challenge, it remains an open question whether it is feasible to discover alternatives to fundamental concepts of RL such as value functions and temporal-difference learning. This paper introduces a new meta-learning approach that discovers an entire update rule which includes both what to predict' (e.g. value functions) andhow to learn from it' (e.g. bootstrapping) by interacting with a set of environments. The output of this method is an RL algorithm that we call Learned Policy Gradient (LPG). Empirical results show that our method discovers its own alternative to the concept of value functions. Furthermore it discovers a bootstrapping mechanism to maintain and use its predictions. Surprisingly, when trained solely on toy environments, LPG generalises effectively to complex Atari games and achieves non-trivial performance. This shows the potential to discover general RL algorithms from data.

Deep reinforcement learning includes a broad family of algorithms that parameterise an internal representation, such as a value function or policy, by a deep neural network. Each algorithm optimises its parameters with respect to an objective, such as Q-learning or policy gradient, that defines its semantics. In this work, we propose an algorithm based on meta-gradient descent that discovers its own objective, flexibly parameterised by a deep neural network, solely from interactive experience with its environment. Over time, this allows the agent to learn how to learn increasingly effectively. Furthermore, because the objective is discovered online, it can adapt to changes over time. We demonstrate that the algorithm discovers how to address several important issues in RL, such as bootstrapping, non-stationarity, and off-policy learning. On the Atari Learning Environment, the meta-gradient algorithm adapts over time to learn with greater efficiency, eventually outperforming the median score of a strong actor-critic baseline.

Reinforcement learning algorithms are highly sensitive to the choice of hyperparameters, typically requiring significant manual effort to identify hyperparameters that perform well on a new domain. In this paper, we take a step towards addressing this issue by using metagradients to automatically adapt hyperparameters online by meta-gradient descent (Xu et al., 2018). We apply our algorithm, Self-Tuning Actor-Critic (STAC), to self-tune all the differentiable hyperparameters of an actor-critic loss function, to discover auxiliary tasks, and to improve off-policy learning using a novel leaky V-trace operator. STAC is simple to use, sample efficient and does not require a significant increase in compute. Ablative studies show that the overall performance of STAC improved as we adapt more hyperparameters. When applied to the Arcade Learning Environment (Bellemare et al. 2012), STAC improved the median human normalized score in 200M steps from 243% to 364%. When applied to the DM Control suite (Tassa et al., 2018), STAC improved the mean score in 30M steps from 217 to 389 when learning with features, from 108 to 202 when learning from pixels, and from 195 to 295 in the Real-World Reinforcement Learning Challenge (Dulac-Arnold et al., 2020).

Arguably, intelligent agents ought to be able to discover their own questions so that in learning answers for them they learn unanticipated useful knowledge and skills; this departs from the focus in much of machine learning on agents learning answers to externally defined questions. We present a novel method for a reinforcement learning (RL) agent to discover questions formulated as general value functions or GVFs, a fairly rich form of knowledge representation. Specifically, our method uses non-myopic meta-gradients to learn GVF-questions such that learning answers to them, as an auxiliary task, induces useful representations for the main task faced by the RL agent. We demonstrate that auxiliary tasks based on the discovered GVFs are sufficient, on their own, to build representations that support main task learning, and that they do so better than popular hand-designed auxiliary tasks from the literature. Furthermore, we show, in the context of Atari2600 videogames, how such auxiliary tasks, meta-learned alongside the main task, can improve the data efficiency of an actor-critic agent.

We aim to significantly reduce the computational cost for classification of temporally untrimmed videos while retaining similar accuracy. Existing video classification methods sample frames with a predefined frequency over entire video. Differently, we propose an end-to-end deep reinforcement approach which enables an agent to classify videos by watching a very small portion of frames like what we do. We make two main contributions. First, information is not equally distributed in video frames along time. An agent needs to watch more carefully when a clip is informative and skip the frames if they are redundant or irrelevant. The proposed approach enables the agent to adapt sampling rate to video content and skip most of the frames without the loss of information. Second, in order to have a confident decision, the number of frames that should be watched by an agent varies greatly from one video to another. We incorporate an adaptive stop network to measure confidence score and generate timely trigger to stop the agent watching videos, which improves efficiency without loss of accuracy. Our approach reduces the computational cost significantly for the large-scale YouTube-8M dataset, while the accuracy remains the same.

The goal of reinforcement learning algorithms is to estimate and/or optimise the value function. However, unlike supervised learning, no teacher or oracle is available to provide the true value function. Instead, the majority of reinforcement learning algorithms estimate and/or optimise a proxy for the value function. This proxy is typically based on a sampled and bootstrapped approximation to the true value function, known as a return. The particular choice of return is one of the chief components determining the nature of the algorithm: the rate at which future rewards are discounted; when and how values should be bootstrapped; or even the nature of the rewards themselves. It is well-known that these decisions are crucial to the overall success of RL algorithms. We discuss a gradient-based meta-learning algorithm that is able to adapt the nature of the return, online, whilst interacting and learning from the environment. When applied to 57 games on the Atari 2600 environment over 200 million frames, our algorithm achieved a new state-of-the-art performance.

Semantic attributes have been increasingly used the past few years for multimedia event detection (MED) with promising results. The motivation is that multimedia events generally consist of lower level components such as objects, scenes, and actions. By characterizing multimedia event videos with semantic attributes, one could exploit more informative cues for improved detection results. Much existing work obtains semantic attributes from images, which may be suboptimal for video analysis since these image-inferred attributes do not carry dynamic information that is essential for videos. To address this issue, we propose to learn semantic attributes from external videos using their semantic labels. We name them video attributes in this paper. In contrast with multimedia event videos, these external videos depict lower level contents such as objects, scenes, and actions. To harness video attributes, we propose an algorithm established on a correlation vector that correlates them to a target event. Consequently, we could incorporate video attributes latently as extra information into the event detector learnt from multimedia event videos in a joint framework. To validate our method, we perform experiments on the real-world large-scale TRECVID MED 2013 and 2014 data sets and compare our method with several state-of-the-art algorithms. The experiments show that our method is advantageous for MED.

With the tremendous advances made by Convolutional Neural Networks (ConvNets) on object recognition, we can now easily obtain adequately reliable machine-labeled annotations easily from predictions by off-the-shelf ConvNets. In this work, we present an abstraction memory based framework for few-shot learning, building upon machine-labeled image annotations. Our method takes large-scale machine-annotated dataset (e.g., OpenImages) as an external memory bank. In the external memory bank, the information is stored in the memory slots in the form of key-value, in which image feature is regarded as the key and the label embedding serves as the value. When queried by the few-shot examples, our model selects visually similar data from the external memory bank and writes the useful information obtained from related external data into another memory bank, i.e., abstraction memory. Long Short-Term Memory (LSTM) controllers and attention mechanisms are utilized to guarantee the data written to the abstraction memory correlates with the query example. The abstraction memory concentrates information from the external memory bank to make the few-shot recognition effective. In the experiments, we first confirm that our model can learn to conduct few-shot object recognition on clean human-labeled data from the ImageNet dataset. Then, we demonstrate that with our model, machine-labeled image annotations are very effective and abundant resources for performing object recognition on novel categories. Experimental results show that our proposed model with machine-labeled annotations achieves great results, with only a 1% difference in accuracy between the machine-labeled annotations and the human-labeled annotations.

Despite the recent success of neural networks in image feature learning, a major problem in the video domain is the lack of sufficient labeled data for learning to model temporal information. In this paper, we propose an unsupervised temporal modeling method that learns from untrimmed videos. The speed of motion varies constantly, e.g., a man may run quickly or slowly. We therefore train a Multirate Visual Recurrent Model (MVRM) by encoding frames of a clip with different intervals. This learning process makes the learned model more capable of dealing with motion speed variance. Given a clip sampled from a video, we use its past and future neighboring clips as the temporal context, and reconstruct the two temporal transitions, i.e., present-past transition and present-future transition, reflecting the temporal information in different views. The proposed method exploits the two transitions simultaneously by incorporating a bidirectional reconstruction which consists of a backward reconstruction and a forward reconstruction. We apply the proposed method to two challenging video tasks, i.e., complex event detection and video captioning, in which it achieves state-of-the-art performance. Notably, our method generates the best single feature for event detection with a relative improvement of 10.4% on the MEDTest-13 dataset and achieves the best performance in video captioning across all evaluation metrics on the YouTube2Text dataset.

Neural networks have a smooth initial inductive bias, such that small changes in input do not lead to large changes in output. However, in reinforcement learning domains with sparse rewards, value functions have non-smooth structure with a characteristic asymmetric discontinuity whenever rewards arrive. We propose a mechanism that learns an interpolation between a direct value estimate and a projected value estimate computed from the encountered reward and the previous estimate. This reduces the need to learn about discontinuities, and thus improves the value function approximation. Furthermore, as the interpolation is learned and state-dependent, our method can deal with heterogeneous observability. We demonstrate that this one change leads to significant improvements on multiple Atari games, when applied to the state-of-the-art A3C algorithm.

Semi-supervised learning is proposed to exploit both labeled and unlabeled data. However, as the scale of data in real world applications increases significantly, conventional semi-supervised algorithms usually lead to massive computational cost and cannot be applied to large scale datasets. In addition, label noise is usually present in the practical applications due to human annotation, which very likely results in remarkable degeneration of performance in semi-supervised methods. To address these two challenges, in this paper, we propose an efficient RObust Semi-Supervised Ensemble Learning (ROSSEL) method, which generates pseudo-labels for unlabeled data using a set of weak annotators, and combines them to approximate the ground-truth labels to assist semi-supervised learning. We formulate the weighted combination process as a multiple label kernel learning (MLKL) problem which can be solved efficiently. Compared with other semi-supervised learning algorithms, the proposed method has linear time complexity. Extensive experiments on five benchmark datasets demonstrate the superior effectiveness, efficiency and robustness of the proposed algorithm.

Recently, deep learning approach, especially deep Convolutional Neural Networks (ConvNets), have achieved overwhelming accuracy with fast processing speed for image classification. Incorporating temporal structure with deep ConvNets for video representation becomes a fundamental problem for video content analysis. In this paper, we propose a new approach, namely Hierarchical Recurrent Neural Encoder (HRNE), to exploit temporal information of videos. Compared to recent video representation inference approaches, this paper makes the following three contributions. First, our HRNE is able to efficiently exploit video temporal structure in a longer range by reducing the length of input information flow, and compositing multiple consecutive inputs at a higher level. Second, computation operations are significantly lessened while attaining more non-linearity. Third, HRNE is able to uncover temporal tran-sitions between frame chunks with different granularities, i.e. it can model the temporal transitions between frames as well as the transitions between segments. We apply the new method to video captioning where temporal information plays a crucial role. Experiments demonstrate that our method outperforms the state-of-the-art on video captioning benchmarks.

In this paper, we propose a discriminative video representation for event detection over a large scale video dataset when only limited hardware resources are available. The focus of this paper is to effectively leverage deep Convolutional Neural Networks (CNNs) to advance event detection, where only frame level static descriptors can be extracted by the existing CNN toolkits. This paper makes two contributions to the inference of CNN video representation. First, while average pooling and max pooling have long been the standard approaches to aggregating frame level static features, we show that performance can be significantly improved by taking advantage of an appropriate encoding method. Second, we propose using a set of latent concept descriptors as the frame descriptor, which enriches visual information while keeping it computationally affordable. The integration of the two contributions results in a new state-of-the-art performance in event detection over the largest video datasets. Compared to improved Dense Trajectories, which has been recognized as the best video representation for event detection, our new representation improves the Mean Average Precision (mAP) from 27.6% to 36.8% for the TRECVID MEDTest 14 dataset and from 34.0% to 44.6% for the TRECVID MEDTest 13 dataset.

Many content-based video search (CBVS) systems have been proposed to analyze the rapidly-increasing amount of user-generated videos on the Internet. Though the accuracy of CBVS systems have drastically improved, these high accuracy systems tend to be too inefficient for interactive search. Therefore, to strive for real-time web-scale CBVS, we perform a comprehensive study on the different components in a CBVS system to understand the trade-offs between accuracy and speed of each component. Directions investigated include exploring different low-level and semantics-based features, testing different compression factors and approximations during video search, and understanding the time v.s. accuracy trade-off of reranking. Extensive experiments on data sets consisting of more than 1,000 hours of video showed that through a combination of effective features, highly compressed representations, and one iteration of reranking, our proposed system can achieve an 10,000-fold speedup while retaining 80% accuracy of a state-of-the-art CBVS system. We further performed search over 1 million videos and demonstrated that our system can complete the search in 0.975 seconds with a single core, which potentially opens the door to interactive web-scale CBVS for the general public.

We address the challenging problem of utilizing related exemplars for complex event detection while multiple features are available. Related exemplars share certain positive elements of the event, but have no uniform pattern due to the huge variance of relevance levels among different related exemplars. None of the existing multiple feature fusion methods can deal with the related exemplars. In this paper, we propose an algorithm which adaptively utilizes the related exemplars by cross-feature learning. Ordinal labels are used to represent the multiple relevance levels of the related videos. Label candidates of related exemplars are generated by exploring the possible relevance levels of each related exemplar via a cross-feature voting strategy. Maximum margin criterion is then applied in our framework to discriminate the positive and negative exemplars, as well as the related exemplars from different relevance levels. We test our algorithm using the large scale TRECVID 2011 dataset and it gains promising performance.

Complex events essentially include human, scenes, objects and actions that can be summarized by visual attributes, so leveraging relevant attributes properly could be helpful for event detection. Many works have exploited attributes at image level for various applications. However, attributes at image level are possibly insufficient for complex event detection in videos due to their limited capability in characterizing the dynamic properties of video data. Hence, we propose to leverage attributes at video level (named as video attributes in this work), i.e., the semantic labels of external videos are used as attributes. Compared to complex event videos, these external videos contain simple contents such as objects, scenes and actions which are the basic elements of complex events. Specifically, building upon a correlation vector which correlates the attributes and the complex event, we incorporate video attributes latently as extra informative cues into the event detector learnt from complex event videos. Extensive experiments on a real-world large-scale dataset validate the efficacy of the proposed approach.

Fusion of multiple features can boost the performance of large-scale visual classification and detection tasks like TRECVID Multimedia Event Detection (MED) competition [1]. In this paper, we propose a novel feature fusion approach, namely Feature Weighting via Optimal Thresholding (FWOT) to effectively fuse various features. FWOT learns the weights, thresholding and smoothing parameters in a joint framework to combine the decision values obtained from all the individual features and the early fusion. To the best of our knowledge, this is the first work to consider the weight and threshold factors of fusion problem simultaneously. Compared to state-of-the-art fusion algorithms, our approach achieves promising improvements on HMDB [8] action recognition dataset and CCV [5] video classification dataset. In addition, experiments on two TRECVID MED 2011 collections show that our approach outperforms the state-of-the-art fusion methods for complex event detection.

Compared to visual concepts such as actions, scenes and objects, complex event is a higher level abstraction of longer video sequences. For example, a "marriage proposal" event is described by multiple objects (e.g., ring, faces), scenes (e.g., in a restaurant, outdoor) and actions (e.g., kneeling down). The positive exemplars which exactly convey the precise semantic of an event are hard to obtain. It would be beneficial to utilize the related exemplars for complex event detection. However, the semantic correlations between related exemplars and the target event vary substantially as relatedness assessment is subjective. Two related exemplars can be about completely different events, e.g., in the TRECVID MED dataset, both bicycle riding and equestrianism are labeled as related to "attempting a bike trick" event. To tackle the subjectiveness of human assessment, our algorithm automatically evaluates how positive the related exemplars are for the detection of an event and uses them on an exemplar-specific basis. Experiments demonstrate that our algorithm is able to utilize related exemplars adaptively, and the algorithm gains good performance for complex event detection.

Multimedia event detection (MED) is an effective technique for video indexing and retrieval. Current classifier training for MED treats the negative videos equally. However, many negative videos may resemble the positive videos in different degrees. Intuitively, we may capture more informative cues from the negative videos if we assign them fine-grained labels, thus benefiting the classifier learning. Aiming for this, we use a statistical method on both the positive and negative examples to get the decisive attributes of a specific event. Based on these decisive attributes, we assign the fine-grained labels to negative examples to treat them differently for more effective exploitation. The resulting fine-grained labels may be not accurate enough to characterize the negative videos. Hence, we propose to jointly optimize the fine-grained labels with the knowledge from the visual features and the attributes representations, which brings mutual reciprocality. Our model obtains two kinds of classifiers, one from the attributes and one from the features, which incorporate the informative cues from the fine-grained labels. The outputs of both classifiers on the testing videos are fused for detection. Extensive experiments on the challenging TRECVID MED 2012 development set have validated the efficacy of our proposed approach.

The selection of discriminative features is an important and effective technique for many multimedia tasks. Using irrelevant features in classification or clustering tasks could deteriorate the performance. Thus, designing efficient feature selection algorithms to remove the irrelevant features is a possible way to improve the classification or clustering performance. With the successful usage of sparse models in image and video classification and understanding, imposing structural sparsity in \emph{feature selection} has been widely investigated during the past years. Motivated by the merit of sparse models, we propose a novel feature selection method using a sparse model in this paper. Different from the state of the art, our method is built upon $\ell _{2,p}$-norm and simultaneously considers both the global and local (GLocal) structures of data distribution. Our method is more flexible in selecting the discriminating features as it is able to control the degree of sparseness. Moreover, considering both global and local structures of data distribution makes our feature selection process more effective. An efficient algorithm is proposed to solve the $\ell_{2,p}$-norm sparsity optimization problem in this paper. Experimental results performed on real-world image and video datasets show the effectiveness of our feature selection method compared to several state-of-the-art methods.