Language Model Guided Interpretable Video Action Reasoning

While neural networks have excelled in video action recognition tasks, their black-box nature often obscures the understanding of their decision-making processes. Recent approaches used inherently interpretable models to analyze video actions in a manner akin to human reasoning. These models, however, usually fall short in performance compared to their black-box counterparts. In this work, we present a new framework named Language-guided Interpretable Action Recognition framework (LaIAR). LaIAR leverages knowledge from language models to enhance both the recognition capabilities and the interpretability of video models. In essence, we redefine the problem of understanding video model decisions as a task of aligning video and language models. Using the logical reasoning captured by the language model, we steer the training of the video model. This integrated approach not only improves the video model's adaptability to different domains but also boosts its overall performance. Extensive experiments on two complex video action datasets, Charades & CAD-120, validates the improved performance and interpretability of our LaIAR framework. The code of LaIAR is available at https://github.com/NingWang2049/LaIAR.

LR-CNN: Lightweight Row-centric Convolutional Neural Network Training for Memory Reduction

In the last decade, Convolutional Neural Network with a multi-layer architecture has advanced rapidly. However, training its complex network is very space-consuming, since a lot of intermediate data are preserved across layers, especially when processing high-dimension inputs with a big batch size. That poses great challenges to the limited memory capacity of current accelerators (e.g., GPUs). Existing efforts mitigate such bottleneck by external auxiliary solutions with additional hardware costs, and internal modifications with potential accuracy penalty. Differently, our analysis reveals that computations intra- and inter-layers exhibit the spatial-temporal weak dependency and even complete independency features. That inspires us to break the traditional layer-by-layer (column) dataflow rule. Now operations are novelly re-organized into rows throughout all convolution layers. This lightweight design allows a majority of intermediate data to be removed without any loss of accuracy. We particularly study the weak dependency between two consecutive rows. For the resulting skewed memory consumption, we give two solutions with different favorite scenarios. Evaluations on two representative networks confirm the effectiveness. We also validate that our middle dataflow optimization can be smoothly embraced by existing works for better memory reduction.

Image classification network enhancement methods based on knowledge injection

The current deep neural network algorithm still stays in the end-to-end training supervision method like Image-Label pairs, which makes traditional algorithm is difficult to explain the reason for the results, and the prediction logic is difficult to understand and analyze. The current algorithm does not use the existing human knowledge information, which makes the model not in line with the human cognition model and makes the model not suitable for human use. In order to solve the above problems, the present invention provides a deep neural network training method based on the human knowledge, which uses the human cognition model to construct the deep neural network training model, and uses the existing human knowledge information to construct the deep neural network training model. This paper proposes a multi-level hierarchical deep learning algorithm, which is composed of multi-level hierarchical deep neural network architecture and multi-level hierarchical deep learning framework. The experimental results show that the proposed algorithm can effectively explain the hidden information of the neural network. The goal of our study is to improve the interpretability of deep neural networks (DNNs) by providing an analysis of the impact of knowledge injection on the classification task. We constructed a knowledge injection dataset with matching knowledge data and image classification data. The knowledge injection dataset is the benchmark dataset for the experiments in the paper. Our model expresses the improvement in interpretability and classification task performance of hidden layers at different scales.

A Video Coding Method Based on Neural Network for CLIC2024

This paper presents a video coding scheme that combines traditional optimization methods with deep learning methods based on the Enhanced Compression Model (ECM). In this paper, the traditional optimization methods adaptively adjust the quantization parameter (QP). The key frame QP offset is set according to the video content characteristics, and the coding tree unit (CTU) level QP of all frames is also adjusted according to the spatial-temporal perception information. Block importance mapping technology (BIM) is also introduced, which adjusts the QP according to the block importance. Meanwhile, the deep learning methods propose a convolutional neural network-based loop filter (CNNLF), which is turned on/off based on the rate-distortion optimization at the CTU and frame level. Besides, intra-prediction using neural networks (NN-intra) is proposed to further improve compression quality, where 8 neural networks are used for predicting blocks of different sizes. The experimental results show that compared with ECM-3.0, the proposed traditional methods and adding deep learning methods improve the PSNR by 0.54 dB and 1 dB at 0.05Mbps, respectively; 0.38 dB and 0.71dB at 0.5 Mbps, respectively, which proves the superiority of our method.

Cycle-Consistency Learning for Captioning and Grounding

We present that visual grounding and image captioning, which perform as two mutually inverse processes, can be bridged together for collaborative training by careful designs. By consolidating this idea, we introduce CyCo, a cyclic-consistent learning framework to ameliorate the independent training pipelines of visual grounding and image captioning. The proposed framework (1) allows the semi-weakly supervised training of visual grounding; (2) improves the performance of fully supervised visual grounding; (3) yields a general captioning model that can describe arbitrary image regions. Extensive experiments show that our fully supervised grounding model achieves state-of-the-art performance, and the semi-weakly supervised one also exhibits competitive performance compared to the fully supervised counterparts. Our image captioning model has the capability to freely describe image regions and meanwhile shows impressive performance on prevalent captioning benchmarks.

DocBinFormer: A Two-Level Transformer Network for Effective Document Image Binarization

In real life, various degradation scenarios exist that might damage document images, making it harder to recognize and analyze them, thus binarization is a fundamental and crucial step for achieving the most optimal performance in any document analysis task. We propose DocBinFormer (Document Binarization Transformer), a novel two-level vision transformer (TL-ViT) architecture based on vision transformers for effective document image binarization. The presented architecture employs a two-level transformer encoder to effectively capture both global and local feature representation from the input images. These complimentary bi-level features are exploited for efficient document image binarization, resulting in improved results for system-generated as well as handwritten document images in a comprehensive approach. With the absence of convolutional layers, the transformer encoder uses the pixel patches and sub-patches along with their positional information to operate directly on them, while the decoder generates a clean (binarized) output image from the latent representation of the patches. Instead of using a simple vision transformer block to extract information from the image patches, the proposed architecture uses two transformer blocks for greater coverage of the extracted feature space on a global and local scale. The encoded feature representation is used by the decoder block to generate the corresponding binarized output. Extensive experiments on a variety of DIBCO and H-DIBCO benchmarks show that the proposed model outperforms state-of-the-art techniques on four metrics. The source code will be made available at https://github.com/RisabBiswas/DocBinFormer.

Throughput Maximization for Intelligent Refracting Surface Assisted mmWave High-Speed Train Communications

With the increasing demands from passengers for data-intensive services, millimeter-wave (mmWave) communication is considered as an effective technique to release the transmission pressure on high speed train (HST) networks. However, mmWave signals ncounter severe losses when passing through the carriage, which decreases the quality of services on board. In this paper, we investigate an intelligent refracting surface (IRS)-assisted HST communication system. Herein, an IRS is deployed on the train window to dynamically reconfigure the propagation environment, and a hybrid time division multiple access-nonorthogonal multiple access scheme is leveraged for interference mitigation. We aim to maximize the overall throughput while taking into account the constraints imposed by base station beamforming, IRS discrete phase shifts and transmit power. To obtain a practical solution, we employ an alternating optimization method and propose a two-stage algorithm. In the first stage, the successive convex approximation method and branch and bound algorithm are leveraged for IRS phase shift design. In the second stage, the Lagrangian multiplier method is utilized for power allocation. Simulation results demonstrate the benefits of IRS adoption and power allocation for throughput improvement in mmWave HST networks.

Scale-MIA: A Scalable Model Inversion Attack against Secure Federated Learning via Latent Space Reconstruction

Federated learning is known for its capability to safeguard participants' data privacy. However, recently emerged model inversion attacks (MIAs) have shown that a malicious parameter server can reconstruct individual users' local data samples through model updates. The state-of-the-art attacks either rely on computation-intensive search-based optimization processes to recover each input batch, making scaling difficult, or they involve the malicious parameter server adding extra modules before the global model architecture, rendering the attacks too conspicuous and easily detectable. To overcome these limitations, we propose Scale-MIA, a novel MIA capable of efficiently and accurately recovering training samples of clients from the aggregated updates, even when the system is under the protection of a robust secure aggregation protocol. Unlike existing approaches treating models as black boxes, Scale-MIA recognizes the importance of the intricate architecture and inner workings of machine learning models. It identifies the latent space as the critical layer for breaching privacy and decomposes the complex recovery task into an innovative two-step process to reduce computation complexity. The first step involves reconstructing the latent space representations (LSRs) from the aggregated model updates using a closed-form inversion mechanism, leveraging specially crafted adversarial linear layers. In the second step, the whole input batches are recovered from the LSRs by feeding them into a fine-tuned generative decoder. We implemented Scale-MIA on multiple commonly used machine learning models and conducted comprehensive experiments across various settings. The results demonstrate that Scale-MIA achieves excellent recovery performance on different datasets, exhibiting high reconstruction rates, accuracy, and attack efficiency on a larger scale compared to state-of-the-art MIAs.

Sum Rate Maximization under AoI Constraints for RIS-Assisted mmWave Communications

The concept of age of information (AoI) has been proposed to quantify information freshness, which is crucial for time-sensitive applications. However, in millimeter wave (mmWave) communication systems, the link blockage caused by obstacles and the severe path loss greatly impair the freshness of information received by the user equipments (UEs). In this paper, we focus on reconfigurable intelligent surface (RIS)-assisted mmWave communications, where beamforming is performed at transceivers to provide directional beam gain and a RIS is deployed to combat link blockage. We aim to maximize the system sum rate while satisfying the information freshness requirements of UEs by jointly optimizing the beamforming at transceivers, the discrete RIS reflection coefficients, and the UE scheduling strategy. To facilitate a practical solution, we decompose the problem into two subproblems. For the first per-UE data rate maximization problem, we further decompose it into a beamforming optimization subproblem and a RIS reflection coefficient optimization subproblem. Considering the difficulty of channel estimation, we utilize the hierarchical search method for the former and the local search method for the latter, and then adopt the block coordinate descent (BCD) method to alternately solve them. For the second scheduling strategy design problem, a low-complexity heuristic scheduling algorithm is designed. Simulation results show that the proposed algorithm can effectively improve the system sum rate while satisfying the information freshness requirements of all UEs.