Vasileios Mezaris

Konstantinos Gkrispanis, Nikolaos Gkalelis, Vasileios Mezaris

Face detectors are becoming a crucial component of many applications, including surveillance, that often have to run on edge devices with limited processing power and memory. Therefore, there's a pressing demand for compact face detection models that can function efficiently across resource-constrained devices. Over recent years, network pruning techniques have attracted a lot of attention from researchers. These methods haven't been well examined in the context of face detectors, despite their expanding popularity. In this paper, we implement filter pruning on two already small and compact face detectors, named EXTD (Extremely Tiny Face Detector) and EResFD (Efficient ResNet Face Detector). The main pruning algorithm that we utilize is Filter Pruning via Geometric Median (FPGM), combined with the Soft Filter Pruning (SFP) iterative procedure. We also apply L1 Norm pruning, as a baseline to compare with the proposed approach. The experimental evaluation on the WIDER FACE dataset indicates that the proposed approach has the potential to further reduce the model size of already lightweight face detectors, with limited accuracy loss, or even with small accuracy gain for low pruning rates.

Maria Tzelepi, Vasileios Mezaris

Memes have become a dominant form of communication in social media in recent years. Memes are typically humorous and harmless, however there are also memes that promote hate speech, being in this way harmful to individuals and groups based on their identity. Therefore, detecting hateful content in memes has emerged as a task of critical importance. The need for understanding the complex interactions of images and their embedded text renders the hateful meme detection a challenging multimodal task. In this paper we propose to address the aforementioned task leveraging knowledge encoded in powerful Large Multimodal Models (LMM). Specifically, we propose to exploit LMMs in a two-fold manner. First, by extracting knowledge oriented to the hateful meme detection task in order to build strong meme representations. Specifically, generic semantic descriptions and emotions that the images along with their embedded texts elicit are extracted, which are then used to train a simple classification head for hateful meme detection. Second, by developing a novel hard mining approach introducing directly LMM-encoded knowledge to the training process, providing further improvements. We perform extensive experiments on two datasets that validate the effectiveness of the proposed method, achieving state-of-the-art performance. Our code and trained models are publicly available at: https://github.com/IDT-ITI/LMM-CLIP-meme.

Nikolaos Kaparinos, Vasileios Mezaris

Face detection is a computer vision application that increasingly demands lightweight models to facilitate deployment on devices with limited computational resources. Neural network pruning is a promising technique that can effectively reduce network size without significantly affecting performance. In this work, we propose a novel face detection pruning pipeline that leverages Filter Pruning via Geometric Median (FPGM) pruning, Soft Filter Pruning (SFP) and Bayesian optimization in order to achieve a superior trade-off between size and performance compared to existing approaches. FPGM pruning is a structured pruning technique that allows pruning the least significant filters in each layer, while SFP iteratively prunes the filters and allows them to be updated in any subsequent training step. Bayesian optimization is employed in order to optimize the pruning rates of each layer, rather than relying on engineering expertise to determine the optimal pruning rates for each layer. In our experiments across all three subsets of the WIDER FACE dataset, our proposed approach B-FPGM consistently outperforms existing ones in balancing model size and performance. All our experiments were applied to EResFD, the currently smallest (in number of parameters) well-performing face detector of the literature; a small ablation study with a second small face detector, EXTD, is also reported. The source code and trained pruned face detection models can be found at: https://github.com/IDTITI/B-FPGM.

Maria Tzelepi, Vasileios Mezaris

In this paper we deal with image classification tasks using the powerful CLIP vision-language model. Our goal is to advance the classification performance using the CLIP's image encoder, by proposing a novel Large Multimodal Model (LMM) based regularization method. The proposed method uses an LMM to extract semantic descriptions for the images of the dataset. Then, it uses the CLIP's text encoder, frozen, in order to obtain the corresponding text embeddings and compute the mean semantic class descriptions. Subsequently, we adapt the CLIP's image encoder by adding a classification head, and we train it along with the image encoder output, apart from the main classification objective, with an additional auxiliary objective. The additional objective forces the embeddings at the image encoder's output to become similar to their corresponding LMM-generated mean semantic class descriptions. In this way, it produces embeddings with enhanced discrimination ability, leading to improved classification performance. The effectiveness of the proposed regularization method is validated through extensive experiments on three image classification datasets.

Nikolaos Kaparinos, Vasileios Mezaris

Generative Adversarial Networks (GANs) achieve excellent performance in generative tasks, such as image super-resolution, but their computational requirements make difficult their deployment on resource-constrained devices. While knowledge distillation is a promising research direction for GAN compression, effectively training a smaller student generator is challenging due to the capacity mismatch between the student generator and the teacher discriminator. In this work, we propose Student Discriminator Assisted Knowledge Distillation (SDAKD), a novel GAN distillation methodology that introduces a student discriminator to mitigate this capacity mismatch. SDAKD follows a three-stage training strategy, and integrates an adapted feature map distillation approach in its last two training stages. We evaluated SDAKD on two well-performing super-resolution GANs, GCFSR and Real-ESRGAN. Our experiments demonstrate consistent improvements over the baselines and SOTA GAN knowledge distillation methods. The SDAKD source code will be made openly available upon acceptance of the paper.

Maria Tzelepi, Vasileios Mezaris

In this paper we deal with the task of Disturbing Image Detection (DID), exploiting knowledge encoded in Large Multimodal Models (LMMs). Specifically, we propose to exploit LMM knowledge in a two-fold manner: first by extracting generic semantic descriptions, and second by extracting elicited emotions. Subsequently, we use the CLIP's text encoder in order to obtain the text embeddings of both the generic semantic descriptions and LMM-elicited emotions. Finally, we use the aforementioned text embeddings along with the corresponding CLIP's image embeddings for performing the DID task. The proposed method significantly improves the baseline classification accuracy, achieving state-of-the-art performance on the augmented Disturbing Image Detection dataset.

Maria Tzelepi, Vasileios Mezaris

In this paper, we aim to improve the performance of a deep learning model towards image classification tasks, proposing a novel anchor-based training methodology, named \textit{Online Anchor-based Training} (OAT). The OAT method, guided by the insights provided in the anchor-based object detection methodologies, instead of learning directly the class labels, proposes to train a model to learn percentage changes of the class labels with respect to defined anchors. We define as anchors the batch centers at the output of the model. Then, during the test phase, the predictions are converted back to the original class label space, and the performance is evaluated. The effectiveness of the OAT method is validated on four datasets.

Maria Tzelepi, Vasileios Mezaris

In this paper we address image classification tasks leveraging knowledge encoded in Large Multimodal Models (LMMs). More specifically, we use the MiniGPT-4 model to extract semantic descriptions for the images, in a multimodal prompting fashion. In the current literature, vision language models such as CLIP, among other approaches, are utilized as feature extractors, using only the image encoder, for solving image classification tasks. In this paper, we propose to additionally use the text encoder to obtain the text embeddings corresponding to the MiniGPT-4-generated semantic descriptions. Thus, we use both the image and text embeddings for solving the image classification task. The experimental evaluation on three datasets validates the improved classification performance achieved by exploiting LMM-based knowledge.