PLPP: Prompt Learning with Perplexity Is Self-Distillation for Vision-Language Models

Pre-trained Vision-Language (VL) models such as CLIP have demonstrated their excellent performance across numerous downstream tasks. A recent method, Context Optimization (CoOp), further improves the performance of VL models on downstream tasks by introducing prompt learning. CoOp optimizes a set of learnable vectors, aka prompt, and freezes the whole CLIP model. However, relying solely on CLIP loss to fine-tune prompts can lead to models that are prone to overfitting on downstream task. To address this issue, we propose a plug-in prompt-regularization method called PLPP (Prompt Learning with PerPlexity), which use perplexity loss to regularize prompt learning. PLPP designs a two-step operation to compute the perplexity for prompts: (a) calculating cosine similarity between the weight of the embedding layer and prompts to get labels, (b) introducing a language model (LM) head that requires no training behind text encoder to output word probability distribution. Meanwhile, we unveil that the essence of PLPP is inherently a form of self-distillation. To further prevent overfitting as well as to reduce the additional computation introduced by PLPP, we turn the hard label to soft label and choose top-$k$ values for calculating the perplexity loss. For accelerating model convergence, we introduce mutual self-distillation learning, that is perplexity and inverted perplexity loss. The experiments conducted on four classification tasks indicate that PLPP exhibits superior performance compared to existing methods.

Bridging the User-side Knowledge Gap in Knowledge-aware Recommendations with Large Language Models

In recent years, knowledge graphs have been integrated into recommender systems as item-side auxiliary information, enhancing recommendation accuracy. However, constructing and integrating structural user-side knowledge remains a significant challenge due to the improper granularity and inherent scarcity of user-side features. Recent advancements in Large Language Models (LLMs) offer the potential to bridge this gap by leveraging their human behavior understanding and extensive real-world knowledge. Nevertheless, integrating LLM-generated information into recommender systems presents challenges, including the risk of noisy information and the need for additional knowledge transfer. In this paper, we propose an LLM-based user-side knowledge inference method alongside a carefully designed recommendation framework to address these challenges. Our approach employs LLMs to infer user interests based on historical behaviors, integrating this user-side information with item-side and collaborative data to construct a hybrid structure: the Collaborative Interest Knowledge Graph (CIKG). Furthermore, we propose a CIKG-based recommendation framework that includes a user interest reconstruction module and a cross-domain contrastive learning module to mitigate potential noise and facilitate knowledge transfer. We conduct extensive experiments on three real-world datasets to validate the effectiveness of our method. Our approach achieves state-of-the-art performance compared to competitive baselines, particularly for users with sparse interactions.

WEPO: Web Element Preference Optimization for LLM-based Web Navigation

The rapid advancement of autonomous web navigation has significantly benefited from grounding pretrained Large Language Models (LLMs) as agents. However, current research has yet to fully leverage the redundancy of HTML elements for contrastive training. This paper introduces a novel approach to LLM-based web navigation tasks, called Web Element Preference Optimization (WEPO). WEPO utilizes unsupervised preference learning by sampling distance-based non-salient web elements as negative samples, optimizing maximum likelihood objective within Direct Preference Optimization (DPO). We evaluate WEPO on the Mind2Web benchmark and empirically demonstrate that WEPO aligns user high-level intent with output actions more effectively. The results show that our method achieved the state-of-the-art, with an improvement of 13.8% over WebAgent and 5.3% over the visual language model CogAgent baseline. Our findings underscore the potential of preference optimization to enhance web navigation and other web page based tasks, suggesting a promising direction for future research.

Desigen: A Pipeline for Controllable Design Template Generation

Templates serve as a good starting point to implement a design (e.g., banner, slide) but it takes great effort from designers to manually create. In this paper, we present Desigen, an automatic template creation pipeline which generates background images as well as harmonious layout elements over the background. Different from natural images, a background image should preserve enough non-salient space for the overlaying layout elements. To equip existing advanced diffusion-based models with stronger spatial control, we propose two simple but effective techniques to constrain the saliency distribution and reduce the attention weight in desired regions during the background generation process. Then conditioned on the background, we synthesize the layout with a Transformer-based autoregressive generator. To achieve a more harmonious composition, we propose an iterative inference strategy to adjust the synthesized background and layout in multiple rounds. We constructed a design dataset with more than 40k advertisement banners to verify our approach. Extensive experiments demonstrate that the proposed pipeline generates high-quality templates comparable to human designers. More than a single-page design, we further show an application of presentation generation that outputs a set of theme-consistent slides. The data and code are available at https://whaohan.github.io/desigen.

Bert4CMR: Cross-Market Recommendation with Bidirectional Encoder Representations from Transformer

Real-world multinational e-commerce companies, such as Amazon and eBay, serve in multiple countries and regions. Obviously, these markets have similar goods but different users. Some markets are data-scarce, while others are data-rich. In recent years, cross-market recommendation (CMR) has been proposed to enhance data-scarce markets by leveraging auxiliary information from data-rich markets. Previous works fine-tune the pre-trained model on the local market after freezing part of the parameters or introducing inter-market similarity into the local market to improve the performance of CMR. However, they generally do not consider eliminating the mutual interference between markets. Therefore, the existing methods are neither unable to learn unbiased general knowledge nor efficient transfer reusable information across markets. In this paper, we propose a novel attention-based model called Bert4CMR to simultaneously improve all markets' recommendation performance. Specifically, we employ the attention mechanism to capture user interests by modelling user behavioural sequences. We pre-train the proposed model on global data to learn the general knowledge of items. Then we fine-tune specific target markets to perform local recommendations. We propose market embedding to model the bias of each market and reduce the mutual inference between the parallel markets. Extensive experiments conducted on seven markets show that our model is state-of-the-art. Our model outperforms the suboptimal model by 4.82%, 4.73%, 7.66% and 6.49% on average of seven datasets in terms of four metrics, respectively. We conduct ablation experiments to analyse the effectiveness of the proposed components. Experimental results indicate that our model is able to learn general knowledge through global data and shield the mutual interference between markets.

Collaborative Recommendation Model Based on Multi-modal Multi-view Attention Network: Movie and literature cases

The existing collaborative recommendation models that use multi-modal information emphasize the representation of users' preferences but easily ignore the representation of users' dislikes. Nevertheless, modelling users' dislikes facilitates comprehensively characterizing user profiles. Thus, the representation of users' dislikes should be integrated into the user modelling when we construct a collaborative recommendation model. In this paper, we propose a novel Collaborative Recommendation Model based on Multi-modal multi-view Attention Network (CRMMAN), in which the users are represented from both preference and dislike views. Specifically, the users' historical interactions are divided into positive and negative interactions, used to model the user's preference and dislike views, respectively. Furthermore, the semantic and structural information extracted from the scene is employed to enrich the item representation. We validate CRMMAN by designing contrast experiments based on two benchmark MovieLens-1M and Book-Crossing datasets. Movielens-1m has about a million ratings, and Book-Crossing has about 300,000 ratings. Compared with the state-of-the-art knowledge-graph-based and multi-modal recommendation methods, the AUC, NDCG@5 and NDCG@10 are improved by 2.08%, 2.20% and 2.26% on average of two datasets. We also conduct controlled experiments to explore the effects of multi-modal information and multi-view mechanism. The experimental results show that both of them enhance the model's performance.

Multi-sensor Suboptimal Fusion Student's $t$ Filter

A multi-sensor fusion Student's $t$ filter is proposed for time-series recursive estimation in the presence of heavy-tailed process and measurement noises. Driven from an information-theoretic optimization, the approach extends the single sensor Student's $t$ Kalman filter based on the suboptimal arithmetic average (AA) fusion approach. To ensure computationally efficient, closed-form $t$ density recursion, reasonable approximation has been used in both local-sensor filtering and inter-sensor fusion calculation. The overall framework accommodates any Gaussian-oriented fusion approach such as the covariance intersection (CI). Simulation demonstrates the effectiveness of the proposed multi-sensor AA fusion-based $t$ filter in dealing with outliers as compared with the classic Gaussian estimator, and the advantage of the AA fusion in comparison with the CI approach and the augmented measurement fusion.

Construct Informative Triplet with Two-stage Hard-sample Generation

In this paper, we propose a robust sample generation scheme to construct informative triplets. The proposed hard sample generation is a two-stage synthesis framework that produces hard samples through effective positive and negative sample generators in two stages, respectively. The first stage stretches the anchor-positive pairs with piecewise linear manipulation and enhances the quality of generated samples by skillfully designing a conditional generative adversarial network to lower the risk of mode collapse. The second stage utilizes an adaptive reverse metric constraint to generate the final hard samples. Extensive experiments on several benchmark datasets verify that our method achieves superior performance than the existing hard-sample generation algorithms. Besides, we also find that our proposed hard sample generation method combining the existing triplet mining strategies can further boost the deep metric learning performance.

Hard-sample Guided Hybrid Contrast Learning for Unsupervised Person Re-Identification

Unsupervised person re-identification (Re-ID) is a promising and very challenging research problem in computer vision. Learning robust and discriminative features with unlabeled data is of central importance to Re-ID. Recently, more attention has been paid to unsupervised Re-ID algorithms based on clustered pseudo-label. However, the previous approaches did not fully exploit information of hard samples, simply using cluster centroid or all instances for contrastive learning. In this paper, we propose a Hard-sample Guided Hybrid Contrast Learning (HHCL) approach combining cluster-level loss with instance-level loss for unsupervised person Re-ID. Our approach applies cluster centroid contrastive loss to ensure that the network is updated in a more stable way. Meanwhile, introduction of a hard instance contrastive loss further mines the discriminative information. Extensive experiments on two popular large-scale Re-ID benchmarks demonstrate that our HHCL outperforms previous state-of-the-art methods and significantly improves the performance of unsupervised person Re-ID. The code of our work is available soon at https://github.com/bupt-ai-cz/HHCL-ReID.

Personal Fixations-Based Object Segmentation with Object Localization and Boundary Preservation

As a natural way for human-computer interaction, fixation provides a promising solution for interactive image segmentation. In this paper, we focus on Personal Fixations-based Object Segmentation (PFOS) to address issues in previous studies, such as the lack of appropriate dataset and the ambiguity in fixations-based interaction. In particular, we first construct a new PFOS dataset by carefully collecting pixel-level binary annotation data over an existing fixation prediction dataset, such dataset is expected to greatly facilitate the study along the line. Then, considering characteristics of personal fixations, we propose a novel network based on Object Localization and Boundary Preservation (OLBP) to segment the gazed objects. Specifically, the OLBP network utilizes an Object Localization Module (OLM) to analyze personal fixations and locates the gazed objects based on the interpretation. Then, a Boundary Preservation Module (BPM) is designed to introduce additional boundary information to guard the completeness of the gazed objects. Moreover, OLBP is organized in the mixed bottom-up and top-down manner with multiple types of deep supervision. Extensive experiments on the constructed PFOS dataset show the superiority of the proposed OLBP network over 17 state-of-the-art methods, and demonstrate the effectiveness of the proposed OLM and BPM components. The constructed PFOS dataset and the proposed OLBP network are available at https://github.com/MathLee/OLBPNet4PFOS.