Real-world object detection systems, such as those in autonomous driving and surveillance, must continuously learn new object categories and simultaneously adapt to changing environmental conditions. Existing approaches, Class Incremental Object Detection (CIOD) and Domain Incremental Object Detection (DIOD) only address one aspect of this challenge. CIOD struggles in unseen domains, while DIOD suffers from catastrophic forgetting when learning new classes, limiting their real-world applicability. To overcome these limitations, we introduce Dual Incremental Object Detection (DuIOD), a more practical setting that simultaneously handles class and domain shifts in an exemplar-free manner. We propose DuET, a Task Arithmetic-based model merging framework that enables stable incremental learning while mitigating sign conflicts through a novel Directional Consistency Loss. Unlike prior methods, DuET is detector-agnostic, allowing models like YOLO11 and RT-DETR to function as real-time incremental object detectors. To comprehensively evaluate both retention and adaptation, we introduce the Retention-Adaptability Index (RAI), which combines the Average Retention Index (Avg RI) for catastrophic forgetting and the Average Generalization Index for domain adaptability into a common ground. Extensive experiments on the Pascal Series and Diverse Weather Series demonstrate DuET's effectiveness, achieving a +13.12% RAI improvement while preserving 89.3% Avg RI on the Pascal Series (4 tasks), as well as a +11.39% RAI improvement with 88.57% Avg RI on the Diverse Weather Series (3 tasks), outperforming existing methods.
In natural language processing (NLP), the focus has shifted from encoder-only tiny language models like BERT to decoder-only large language models(LLMs) such as GPT-3. However, LLMs' practical application in the financial sector has revealed three limitations: (1) LLMs often perform worse than fine-tuned BERT on discriminative tasks despite costing much higher computational resources, such as market sentiment analysis in financial reports; (2) Application on generative tasks heavily relies on retrieval augmented generation (RAG) methods to provide current and specialized information, with general retrievers showing suboptimal performance on domain-specific retrieval tasks; (3) There are additional inadequacies in other feature-based scenarios, such as topic modeling. We introduce FinBERT2, a specialized bidirectional encoder pretrained on a high-quality, financial-specific corpus of 32b tokens. This represents the largest known Chinese financial pretraining corpus for models of this parameter size. As a better backbone, FinBERT2 can bridge the gap in the financial-specific deployment of LLMs through the following achievements: (1) Discriminative fine-tuned models (Fin-Labelers) outperform other (Fin)BERT variants by 0.4%-3.3% and leading LLMs by 9.7%-12.3% on average across five financial classification tasks. (2) Contrastive fine-tuned models (Fin-Retrievers) outperform both open-source (e.g., +6.8\% avg improvement over BGE-base-zh) and proprietary (e.g., +4.2\% avg improvement over OpenAI's text-embedding-3-large) embedders across five financial retrieval tasks; (3) Building on FinBERT2 variants, we construct the Fin-TopicModel, which enables superior clustering and topic representation for financial titles. Our work revisits financial BERT models through comparative analysis with contemporary LLMs and offers practical insights for effectively utilizing FinBERT in the LLMs era.
Although long-video understanding demands that models capture hierarchical temporal information -- from clip (seconds) and shot (tens of seconds) to event (minutes) and story (hours) -- existing benchmarks either neglect this multi-scale design or scatter scale-specific questions across different videos, preventing direct comparison of model performance across timescales on the same content. To address this, we introduce ScaleLong, the first benchmark to disentangle these factors by embedding questions targeting four hierarchical timescales -- clip (seconds), shot (tens of seconds), event (minutes), and story (hours) -- all within the same video content. This within-content multi-timescale questioning design enables direct comparison of model performance across timescales on identical videos. ScaleLong features 269 long videos (avg.\ 86\,min) from 5 main categories and 36 sub-categories, with 4--8 carefully designed questions, including at least one question for each timescale. Evaluating 23 MLLMs reveals a U-shaped performance curve, with higher accuracy at the shortest and longest timescales and a dip at intermediate levels. Furthermore, ablation studies show that increased visual token capacity consistently enhances reasoning across all timescales. ScaleLong offers a fine-grained, multi-timescale benchmark for advancing MLLM capabilities in long-video understanding. The code and dataset are available https://github.com/multimodal-art-projection/ScaleLong.
Deep models, such as convolutional neural networks (CNNs) and vision transformer (ViT), demonstrate remarkable performance in image classification. However, those deep models require large data to fine-tune, which is impractical in the medical domain due to the data privacy issue. Furthermore, despite the feasible performance of contrastive language image pre-training (CLIP) in the natural domain, the potential of CLIP has not been fully investigated in the medical field. To face these challenges, we considered three scenarios: 1) we introduce a novel CLIP variant using four CNNs and eight ViTs as image encoders for the classification of brain cancer and skin cancer, 2) we combine 12 deep models with two federated learning techniques to protect data privacy, and 3) we involve traditional machine learning (ML) methods to improve the generalization ability of those deep models in unseen domain data. The experimental results indicate that maxvit shows the highest averaged (AVG) test metrics (AVG = 87.03\%) in HAM10000 dataset with multimodal learning, while convnext\_l demonstrates remarkable test with an F1-score of 83.98\% compared to swin\_b with 81.33\% in FL model. Furthermore, the use of support vector machine (SVM) can improve the overall test metrics with AVG of $\sim 2\%$ for swin transformer series in ISIC2018. Our codes are available at https://github.com/AIPMLab/SkinCancerSimulation.
Ultra-high-resolution (UHR) remote sensing (RS) imagery offers valuable data for Earth observation but pose challenges for existing multimodal foundation models due to two key bottlenecks: (1) limited availability of UHR training data, and (2) token explosion caused by the large image size. To address data scarcity, we introduce SuperRS-VQA (avg. 8,376$\times$8,376) and HighRS-VQA (avg. 2,000$\times$1,912), the highest-resolution vision-language datasets in RS to date, covering 22 real-world dialogue tasks. To mitigate token explosion, our pilot studies reveal significant redundancy in RS images: crucial information is concentrated in a small subset of object-centric tokens, while pruning background tokens (e.g., ocean or forest) can even improve performance. Motivated by these findings, we propose two strategies: Background Token Pruning and Anchored Token Selection, to reduce the memory footprint while preserving key semantics.Integrating these techniques, we introduce GeoLLaVA-8K, the first RS-focused multimodal large language model capable of handling inputs up to 8K$\times$8K resolution, built on the LLaVA framework. Trained on SuperRS-VQA and HighRS-VQA, GeoLLaVA-8K sets a new state-of-the-art on the XLRS-Bench.
Spiking Transformers offer an energy-efficient alternative to conventional deep learning by transmitting information solely through binary (0/1) spikes. However, there remains a substantial performance gap compared to artificial neural networks. A common belief is that their binary and sparse activation transmission leads to information loss, thus degrading feature representation and accuracy. In this work, however, we reveal for the first time that spiking neurons preferentially propagate low-frequency information. We hypothesize that the rapid dissipation of high-frequency components is the primary cause of performance degradation. For example, on Cifar-100, adopting Avg-Pooling (low-pass) for token mixing lowers performance to 76.73%; interestingly, replacing it with Max-Pooling (high-pass) pushes the top-1 accuracy to 79.12%, surpassing the well-tuned Spikformer baseline by 0.97%. Accordingly, we introduce Max-Former that restores high-frequency signals through two frequency-enhancing operators: extra Max-Pooling in patch embedding and Depth-Wise Convolution in place of self-attention. Notably, our Max-Former (63.99 M) hits the top-1 accuracy of 82.39% on ImageNet, showing a +7.58% improvement over Spikformer with comparable model size (74.81%, 66.34 M). We hope this simple yet effective solution inspires future research to explore the distinctive nature of spiking neural networks, beyond the established practice in standard deep learning.
Document Layout Analysis (DLA) is a fundamental task in document understanding. However, existing DLA and adaptation methods often require access to large-scale source data and target labels. This requirements severely limiting their real-world applicability, particularly in privacy-sensitive and resource-constrained domains, such as financial statements, medical records, and proprietary business documents. According to our observation, directly transferring source-domain fine-tuned models on target domains often results in a significant performance drop (Avg. -32.64%). In this work, we introduce Source-Free Document Layout Analysis (SFDLA), aiming for adapting a pre-trained source DLA models to an unlabeled target domain, without access to any source data. To address this challenge, we establish the first SFDLA benchmark, covering three major DLA datasets for geometric- and content-aware adaptation. Furthermore, we propose Document Layout Analysis Adapter (DLAdapter), a novel framework that is designed to improve source-free adaptation across document domains. Our method achieves a +4.21% improvement over the source-only baseline and a +2.26% gain over existing source-free methods from PubLayNet to DocLayNet. We believe this work will inspire the DLA community to further investigate source-free document understanding. To support future research of the community, the benchmark, models, and code will be publicly available at https://github.com/s3setewe/sfdla-DLAdapter.
Coresets have become an invaluable tool for solving $k$-means and kernel $k$-means clustering problems on large datasets with small numbers of clusters. On the other hand, spectral clustering works well on sparse graphs and has recently been extended to scale efficiently to large numbers of clusters. We exploit the connection between kernel $k$-means and the normalised cut problem to combine the benefits of both. Our main result is a coreset spectral clustering algorithm for graphs that clusters a coreset graph to infer a good labelling of the original graph. We prove that an $\alpha$-approximation for the normalised cut problem on the coreset graph is an $O(\alpha)$-approximation on the original. We also improve the running time of the state-of-the-art coreset algorithm for kernel $k$-means on sparse kernels, from $\tilde{O}(nk)$ to $\tilde{O}(n\cdot \min \{k, d_{avg}\})$, where $d_{avg}$ is the average number of non-zero entries in each row of the $n\times n$ kernel matrix. Our experiments confirm our coreset algorithm is asymptotically faster on large real-world graphs with many clusters, and show that our clustering algorithm overcomes the main challenge faced by coreset kernel $k$-means on sparse kernels which is getting stuck in local optima.
We propose Partition Dimensions Across (PDX), a data layout for vectors (e.g., embeddings) that, similar to PAX [6], stores multiple vectors in one block, using a vertical layout for the dimensions (Figure 1). PDX accelerates exact and approximate similarity search thanks to its dimension-by-dimension search strategy that operates on multiple-vectors-at-a-time in tight loops. It beats SIMD-optimized distance kernels on standard horizontal vector storage (avg 40% faster), only relying on scalar code that gets auto-vectorized. We combined the PDX layout with recent dimension-pruning algorithms ADSampling [19] and BSA [52] that accelerate approximate vector search. We found that these algorithms on the horizontal vector layout can lose to SIMD-optimized linear scans, even if they are SIMD-optimized. However, when used on PDX, their benefit is restored to 2-7x. We find that search on PDX is especially fast if a limited number of dimensions has to be scanned fully, which is what the dimension-pruning approaches do. We finally introduce PDX-BOND, an even more flexible dimension-pruning strategy, with good performance on exact search and reasonable performance on approximate search. Unlike previous pruning algorithms, it can work on vector data "as-is" without preprocessing; making it attractive for vector databases with frequent updates.
Off-policy policy evaluation (OPE), an essential component of reinforcement learning, has long suffered from stationary state distribution mismatch, undermining both stability and accuracy of OPE estimates. While existing methods correct distribution shifts by estimating density ratios, they often rely on expensive optimization or backward Bellman-based updates and struggle to outperform simpler baselines. We introduce AVG-DICE, a computationally simple Monte Carlo estimator for the density ratio that averages discounted importance sampling ratios, providing an unbiased and consistent correction. AVG-DICE extends naturally to nonlinear function approximation using regression, which we roughly tune and test on OPE tasks based on Mujoco Gym environments and compare with state-of-the-art density-ratio estimators using their reported hyperparameters. In our experiments, AVG-DICE is at least as accurate as state-of-the-art estimators and sometimes offers orders-of-magnitude improvements. However, a sensitivity analysis shows that best-performing hyperparameters may vary substantially across different discount factors, so a re-tuning is suggested.
Video Moment Retrieval is a common task to evaluate the performance of visual-language models - it involves localising start and end times of moments in videos from query sentences. The current task formulation assumes that the queried moment is present in the video, resulting in false positive moment predictions when irrelevant query sentences are provided. In this paper we propose the task of Negative-Aware Video Moment Retrieval (NA-VMR), which considers both moment retrieval accuracy and negative query rejection accuracy. We make the distinction between In-Domain and Out-of-Domain negative queries and provide new evaluation benchmarks for two popular video moment retrieval datasets: QVHighlights and Charades-STA. We analyse the ability of current SOTA video moment retrieval approaches to adapt to Negative-Aware Video Moment Retrieval and propose UniVTG-NA, an adaptation of UniVTG designed to tackle NA-VMR. UniVTG-NA achieves high negative rejection accuracy (avg. $98.4\%$) scores while retaining moment retrieval scores to within $3.87\%$ Recall@1. Dataset splits and code are available at https://github.com/keflanagan/MomentofUntruth
A vast amount of medical knowledge is available for public use through online health forums, and question-answering platforms on social media. The majority of the population in the United States doesn't have the right amount of health literacy to make the best use of that information. Health literacy means the ability to obtain and comprehend the basic health information to make appropriate health decisions. To build the bridge between this gap, organizations advocate adapting this medical knowledge into plain language. Building robust systems to automate the adaptations helps both medical and non-medical professionals best leverage the available information online. The goal of the Plain Language Adaptation of Biomedical Abstracts (PLABA) track is to adapt the biomedical abstracts in English language extracted from PubMed based on the questions asked in MedlinePlus for the general public using plain language at the sentence level. As part of this track, we leveraged the best open-source Large Language Models suitable and fine-tuned for dialog use cases. We compare and present the results for all of our systems and our ranking among the other participants' submissions. Our top performing GPT-4 based model ranked first in the avg. simplicity measure and 3rd on the avg. accuracy measure.
Aquaculture involves cultivating marine and freshwater organisms, with real-time monitoring of aquatic parameters being crucial in fish farming. This thesis proposes an IoT-based framework using sensors and Arduino for efficient monitoring and control of water quality. Different sensors including pH, temperature, and turbidity are placed in cultivating pond water and each of them is connected to a common microcontroller board built on an Arduino Uno. The sensors read the data from the water and store it as a CSV file in an IoT cloud named Thingspeak through the Arduino Microcontroller. In the experimental part, we collected data from 5 ponds with various sizes and environments. After getting the real-time data, we compared these with the standard reference values. As a result, we can make the decision about which ponds are satisfactory for cultivating fish and what is not. After that, we labeled the data with 11 fish categories including Katla, sing, prawn, rui, koi, pangas, tilapia, silvercarp, karpio, magur, and shrimp. In addition, the data were analyzed using 10 machine learning (ML) algorithms containing J48, Random Forest, K-NN, K*, LMT, REPTree, JRIP, PART, Decision Table, and Logit boost. After experimental evaluation, it was observed among 5 ponds, only three ponds were perfect for fish farming, where these 3 ponds only satisfied the standard reference values of pH (6.5-8.5), Temperature (16-24)oC, Turbidity (below 10)ntu, Conductivity (970-1825){\mu}S/cm, and Depth (1-4) meter. Among the state-of-the-art machine learning algorithms, Random Forest achieved the highest score of performance metrics as accuracy 94.42%, kappa statistics 93.5%, and Avg. TP Rate 94.4%. In addition, we calculated the BOD, COD, and DO for one scenario. This study includes details of the proposed IoT system's prototype hardware.
Given a video and a linguistic query, video moment retrieval and highlight detection (MR&HD) aim to locate all the relevant spans while simultaneously predicting saliency scores. Most existing methods utilize RGB images as input, overlooking the inherent multi-modal visual signals like optical flow and depth. In this paper, we propose a Multi-modal Fusion and Query Refinement Network (MRNet) to learn complementary information from multi-modal cues. Specifically, we design a multi-modal fusion module to dynamically combine RGB, optical flow, and depth map. Furthermore, to simulate human understanding of sentences, we introduce a query refinement module that merges text at different granularities, containing word-, phrase-, and sentence-wise levels. Comprehensive experiments on QVHighlights and Charades datasets indicate that MRNet outperforms current state-of-the-art methods, achieving notable improvements in MR-mAP@Avg (+3.41) and HD-HIT@1 (+3.46) on QVHighlights.
The use of eXplainable Artificial Intelligence (XAI) systems has introduced a set of challenges that need resolution. The XAI robustness, or stability, has been one of the goals of the community from its beginning. Multiple authors have proposed evaluating this feature using objective evaluation measures. Nonetheless, many questions remain. With this work, we propose a novel approach to meta-evaluate these metrics, i.e. analyze the correctness of the evaluators. We propose two new tests that allowed us to evaluate two different stability measures: AVG-Sensitiviy and MAX-Senstivity. We tested their reliability in the presence of perfect and robust explanations, generated with a Decision Tree; as well as completely random explanations and prediction. The metrics results showed their incapacity of identify as erroneous the random explanations, highlighting their overall unreliability.
In the ever-evolving landscape of natural language processing and information retrieval, the need for robust and domain-specific entity linking algorithms has become increasingly apparent. It is crucial in a considerable number of fields such as humanities, technical writing and biomedical sciences to enrich texts with semantics and discover more knowledge. The use of Named Entity Disambiguation (NED) in such domains requires handling noisy texts, low resource settings and domain-specific KBs. Existing approaches are mostly inappropriate for such scenarios, as they either depend on training data or are not flexible enough to work with domain-specific KBs. Thus in this work, we present an unsupervised approach leveraging the concept of Group Steiner Trees (GST), which can identify the most relevant candidates for entity disambiguation using the contextual similarities across candidate entities for all the mentions present in a document. We outperform the state-of-the-art unsupervised methods by more than 40\% (in avg.) in terms of Precision@1 across various domain-specific datasets.
We introduce EgoPoints, a benchmark for point tracking in egocentric videos. We annotate 4.7K challenging tracks in egocentric sequences. Compared to the popular TAP-Vid-DAVIS evaluation benchmark, we include 9x more points that go out-of-view and 59x more points that require re-identification (ReID) after returning to view. To measure the performance of models on these challenging points, we introduce evaluation metrics that specifically monitor tracking performance on points in-view, out-of-view, and points that require re-identification. We then propose a pipeline to create semi-real sequences, with automatic ground truth. We generate 11K such sequences by combining dynamic Kubric objects with scene points from EPIC Fields. When fine-tuning point tracking methods on these sequences and evaluating on our annotated EgoPoints sequences, we improve CoTracker across all metrics, including the tracking accuracy $\delta^\star_{\text{avg}}$ by 2.7 percentage points and accuracy on ReID sequences (ReID$\delta_{\text{avg}}$) by 2.4 points. We also improve $\delta^\star_{\text{avg}}$ and ReID$\delta_{\text{avg}}$ of PIPs++ by 0.3 and 2.8 respectively.
Deep generative models have achieved tremendous success in structure-based drug design in recent years, especially for generating 3D ligand molecules that bind to specific protein pocket. Notably, diffusion models have transformed ligand generation by providing exceptional quality and creativity. However, traditional diffusion models are restricted by their conventional learning objectives, which limit their broader applicability. In this work, we propose a new framework FlowSBDD, which is based on rectified flow model, allows us to flexibly incorporate additional loss to optimize specific target and introduce additional condition either as an extra input condition or replacing the initial Gaussian distribution. Extensive experiments on CrossDocked2020 show that our approach could achieve state-of-the-art performance on generating high-affinity molecules while maintaining proper molecular properties without specifically designing binding site, with up to -8.50 Avg. Vina Dock score and 75.0% Diversity.
To reduce the latency associated with autoretrogressive LLM inference, speculative decoding has emerged as a novel decoding paradigm, where future tokens are drafted and verified in parallel. However, the practical deployment of speculative decoding is hindered by its requirements for additional computational resources and fine-tuning, which limits its out-of-the-box usability. To address these challenges, we present PLD+, a suite of novel algorithms developed to accelerate the inference process of LLMs, particularly for input-guided tasks. These tasks, which include code editing, text editing, summarization, etc., often feature outputs with substantial overlap with their inputs-an attribute PLD+ is designed to exploit. PLD+ also leverages the artifacts (attention and hidden states) generated during inference to accelerate inference speed. We test our approach on five input-guided tasks and through extensive experiments we find that PLD+ outperforms all tuning-free approaches. In the greedy setting, it even outperforms the state-of-the-art tuning-dependent approach EAGLE on four of the tasks. (by a margin of upto 2.31 in terms of avg. speedup). Our approach is tuning free, does not require any additional compute and can easily be used for accelerating inference of any LLM.
Modern deep policy gradient methods achieve effective performance on simulated robotic tasks, but they all require large replay buffers or expensive batch updates, or both, making them incompatible for real systems with resource-limited computers. We show that these methods fail catastrophically when limited to small replay buffers or during incremental learning, where updates only use the most recent sample without batch updates or a replay buffer. We propose a novel incremental deep policy gradient method -- Action Value Gradient (AVG) and a set of normalization and scaling techniques to address the challenges of instability in incremental learning. On robotic simulation benchmarks, we show that AVG is the only incremental method that learns effectively, often achieving final performance comparable to batch policy gradient methods. This advancement enabled us to show for the first time effective deep reinforcement learning with real robots using only incremental updates, employing a robotic manipulator and a mobile robot.