Search

High-speed tactile arrays are essential for real-time robotic control in unstructured environments, but high pixel counts limit readout rates of most large tactile arrays to below 100Hz. We introduce ACTS - adaptive compressive tactile subsampling - a method that efficiently samples tactile matrices and reconstructs interactions using sparse recovery and a learned tactile dictionary. Tested on a 1024-pixel sensor array (32x32), ACTS increased frame rates by 18X compared to raster scanning, with minimal error. For the first time in large-area tactile skin, we demonstrate rapid object classification within 20ms of contact, high-speed projectile detection, ricochet angle estimation, and deformation tracking through enhanced spatiotemporal resolution. Our method can be implemented in firmware, upgrading existing low-cost, flexible, and robust tactile arrays into high-resolution systems for large-area spatiotemporal touch sensing., Comment: 40 pages, 8 main figures, 12 supplemental figures, Videos can be accessed at https://tinyurl.com/TactileSubsampling

Humor is a culturally nuanced aspect of human language that presents challenges for understanding and generation, requiring participants to possess good creativity and strong associative thinking. Similar to reasoning tasks like solving math problems, humor generation requires continuous reflection and revision to foster creative thinking, rather than relying on a sudden flash of inspiration like Creative Leap-of-Thought (CLoT) paradigm. Although CLoT can realize the ability of remote association generation, this paradigm fails to generate humor content. Therefore, in this paper, we propose a systematic way of thinking about generating humor and based on it, we built Creative Leap of Structured Thought (CLoST) frame. First, a reward model is necessary achieve the purpose of being able to correct errors, since there is currently no expert model of humor and a usable rule to determine whether a piece of content is humorous. Judgement-oriented instructions are designed to improve the capability of a model, and we also propose an open-domain instruction evolutionary method to fully unleash the potential. Then, through reinforcement learning, the model learns to hone its rationales of the thought chain and refine the strategies it uses. Thus, it learns to recognize and correct its mistakes, and finally generate the most humorous and creative answer. These findings deepen our understanding of the creative capabilities of LLMs and provide ways to enhance LLMs' creative abilities for cross-domain innovative applications.

The rapid development in the performance of large language models (LLMs) is accompanied by the escalation of model size, leading to the increasing cost of model training and inference. Previous research has discovered that certain layers in LLMs exhibit redundancy, and removing these layers brings only marginal loss in model performance. In this paper, we adopt the probing technique to explain the layer redundancy in LLMs and demonstrate that language models can be effectively pruned with probing classifiers. We propose chip-tuning, a simple and effective structured pruning framework specialized for classification problems. Chip-tuning attaches tiny probing classifiers named chips to different layers of LLMs, and trains chips with the backbone model frozen. After selecting a chip for classification, all layers subsequent to the attached layer could be removed with marginal performance loss. Experimental results on various LLMs and datasets demonstrate that chip-tuning significantly outperforms previous state-of-the-art baselines in both accuracy and pruning ratio, achieving a pruning ratio of up to 50%. We also find that chip-tuning could be applied on multimodal models, and could be combined with model finetuning, proving its excellent compatibility.

Multi-modal large language models (MLLMs) have demonstrated considerable potential across various downstream tasks that require cross-domain knowledge. MLLMs capable of processing videos, known as Video-MLLMs, have attracted broad interest in video-language understanding. However, videos, especially long videos, contain more visual tokens than images, making them difficult for LLMs to process. Existing works either downsample visual features or extend the LLM context size, risking the loss of high-resolution information or slowing down inference speed. To address these limitations, we apply cross-attention layers in the intermediate projector between the visual encoder and the large language model (LLM). As the naive cross-attention mechanism is insensitive to temporal order, we further introduce causal cross-attention masks (CCAMs) within the cross-attention layers. This Video-MLLM, named Video-CCAM, is trained in a straightforward two-stage fashion: feature alignment and visual instruction tuning. We develop several Video-CCAM models based on LLMs of different sizes (4B, 9B, and 14B). Video-CCAM proves to be a robust Video-MLLM and shows outstanding performance from short videos to long ones. Among standard video benchmarks like MVBench and VideoChatGPT-QA, Video-CCAM shows outstanding performances (1st/2nd/3rd in MVBench and TGIF-QA, 2nd/3rd/4th in MSVD-QA, MSRVTT-QA, and ActivityNet-QA). In benchmarks encompassing long videos, Video-CCAM models can be directly adapted to long video understanding and still achieve exceptional scores despite being trained solely with images and 16-frame videos. Using 96 frames (6$\times$ the training number of frames), Video-CCAM models rank 1st/2nd/3rd in VideoVista and 1st/2nd/4th in MLVU among all open-source Video-MLLMs, respectively. The code is publicly available in \url{https://github.com/QQ-MM/Video-CCAM}., Comment: 10 pages, 5 figures

Utilizing tools with Large Language Models (LLMs) is essential for grounding AI agents in real-world applications. The prevailing approach involves few-shot prompting with demonstrations or fine-tuning with expert annotations. However, mere in-context demonstrations may fail to cover sufficient knowledge for complex tools and tasks. Training on solution paths is also hindered by the high cost of expert annotations and generalizing to new tools. A core challenge of generalizable tool use lies in understanding the "meta", or fundamental natures of tools that are transferable across tasks, such as causality and constraints. In this paper, we present MetaTool, a novel tool learning methodology designed to generalize across any reusable toolset. Our approach incorporates a self-supervised augmentation technique derived from a series of meta-tasks. This involves predicting masked elements in the tool execution process. The self-supervised procedure enables scalable generation of high-quality QA data, which is handy for supervising tool understanding. By incorporating meta-task data into task-oriented training, our method significantly enhances the performance of open-source LLMs, achieving results comparable to ChatGPT in both tool-based planning and chatting scenarios. Through large-scale instruction tuning, the MetaTool model demonstrates impressive zero-shot generalizability on new tasks., Comment: 9 pages, 4 figures

Whether diffuse mismatch model for electrons (DMMe) hold true in more general cases remains largely unexplored, especially in cases where at least one material does not behave like a free-electron metal and/or the interface is smooth enough to allow non-diffuse transmission of electrons. In this study, DMMe was proposed to predict the thermal conductance of metal-metal interfaces. A set of aluminum-X samples (X = Cu, Ag, Fe, Ni) were grown and the time domain thermoreflectance (TDTR) technique was used to measure the metal-metal interface conductance. It was then compared to the two variants of the DMMe-using both a crude theory based on free-electron metals and accurate band structures provided by density functional theory., Comment: 34 Pages, 15 Figures

In the previous studies, the ultraviolet light thermoreflectance (UV-TDTR) signal from bulk semiconducting samples cannot be well explained by a thermal models based on the assumption that the heat is both absorbed and probed near the surface. A thermoreflectance (TDTR) technique was developed to directly excite semiconductors using UV-TDTR. At {\lambda} = 400nm, the photon energy is much greater than most semiconducting bandgaps, potentially allowing semiconducting transducers to absorb light within about 10 nm of the surface, potentially enabling direct measurements of semiconductor-semiconductor interfaces. The thermoreflectance coefficient for some direct-bandgap semiconducting materials was measured. Thermal transport models were fitted to thermorefectance data collected from the system. The differences in the signal for semiconductors with different doping and nanostructure features were also studied, which should change the recombination rate. No significant changes to the signal were observed., Comment: 20 Pages, 5 Figures

Instruction tuning is an essential supervised training phase for Large Language Models (LLMs), with the goal of enhancing LLMs' capacity to generalize instruction execution and adapt to user preferences. With the growing incorporation of multi-modal data into LLMs, there is an increasing interest in the performance of vision-language instruction tuning which presents more complex features in comparison to pure text instructions. In this paper, we systematically review the latest vision-language instruction tuning settings and datasets in multi-modal LLMs and summarize the characteristics that high-quality vision-language tuning data should have. We consider these characteristics as the foundational principles for constructing vision-language instruction data and propose a complete construction pipeline consisting of data collection, instruction generation, and quality control modules that incorporate meticulously designed instruction property evaluation indicators. We perform vision-language instruction tuning on three widely used multi-modal LLMs based on the instruction data we constructed and conduct extensive experiments on the corresponding metrics to demonstrate the rationality of the construction principles proposed in this paper. The code and dataset related to this paper have been open-sourced at \url{https://github.com/palchenli/VL-Instruction-Tuning}., Comment: 36 pages, 6 figures

This paper introduces the HumTrans dataset, which is publicly available and primarily designed for humming melody transcription. The dataset can also serve as a foundation for downstream tasks such as humming melody based music generation. It consists of 500 musical compositions of different genres and languages, with each composition divided into multiple segments. In total, the dataset comprises 1000 music segments. To collect this humming dataset, we employed 10 college students, all of whom are either music majors or proficient in playing at least one musical instrument. Each of them hummed every segment twice using the web recording interface provided by our designed website. The humming recordings were sampled at a frequency of 44,100 Hz. During the humming session, the main interface provides a musical score for students to reference, with the melody audio playing simultaneously to aid in capturing both melody and rhythm. The dataset encompasses approximately 56.22 hours of audio, making it the largest known humming dataset to date. The dataset will be released on Hugging Face, and we will provide a GitHub repository containing baseline results and evaluation codes.

Background music (BGM) can enhance the video's emotion. However, selecting an appropriate BGM often requires domain knowledge. This has led to the development of video-music retrieval techniques. Most existing approaches utilize pretrained video/music feature extractors trained with different target sets to obtain average video/music-level embeddings. The drawbacks are two-fold. One is that different target sets for video/music pretraining may cause the generated embeddings difficult to match. The second is that the underlying temporal correlation between video and music is ignored. In this paper, our proposed approach leverages a unified target set to perform video/music pretraining and produces clip-level embeddings to preserve temporal information. The downstream cross-modal matching is based on the clip-level features with embedded music rhythm and optical flow information. Experiments demonstrate that our proposed method can achieve superior performance over the state-of-the-art methods by a significant margin.

Tags are pivotal in facilitating the effective distribution of multimedia content in various applications in the contemporary Internet era, such as search engines and recommendation systems. Recently, large language models (LLMs) have demonstrated impressive capabilities across a wide range of tasks. In this work, we propose TagGPT, a fully automated system capable of tag extraction and multimodal tagging in a completely zero-shot fashion. Our core insight is that, through elaborate prompt engineering, LLMs are able to extract and reason about proper tags given textual clues of multimodal data, e.g., OCR, ASR, title, etc. Specifically, to automatically build a high-quality tag set that reflects user intent and interests for a specific application, TagGPT predicts large-scale candidate tags from a series of raw data via prompting LLMs, filtered with frequency and semantics. Given a new entity that needs tagging for distribution, TagGPT introduces two alternative options for zero-shot tagging, i.e., a generative method with late semantic matching with the tag set, and another selective method with early matching in prompts. It is well noticed that TagGPT provides a system-level solution based on a modular framework equipped with a pre-trained LLM (GPT-3.5 used here) and a sentence embedding model (SimCSE used here), which can be seamlessly replaced with any more advanced one you want. TagGPT is applicable for various modalities of data in modern social media and showcases strong generalization ability to a wide range of applications. We evaluate TagGPT on publicly available datasets, i.e., Kuaishou and Food.com, and demonstrate the effectiveness of TagGPT compared to existing hashtags and off-the-shelf taggers. Project page: https://github.com/TencentARC/TagGPT., Comment: 13 pages, 6 figures

Both masked image modeling (MIM) and natural language supervision have facilitated the progress of transferable visual pre-training. In this work, we seek the synergy between two paradigms and study the emerging properties when MIM meets natural language supervision. To this end, we present a novel masked visual Reconstruction In Language semantic Space (RILS) pre-training framework, in which sentence representations, encoded by the text encoder, serve as prototypes to transform the vision-only signals into patch-sentence probabilities as semantically meaningful MIM reconstruction targets. The vision models can therefore capture useful components with structured information by predicting proper semantic of masked tokens. Better visual representations could, in turn, improve the text encoder via the image-text alignment objective, which is essential for the effective MIM target transformation. Extensive experimental results demonstrate that our method not only enjoys the best of previous MIM and CLIP but also achieves further improvements on various tasks due to their mutual benefits. RILS exhibits advanced transferability on downstream classification, detection, and segmentation, especially for low-shot regimes. Code will be made available at https://github.com/hustvl/RILS.

Since the development of self-supervised visual representation learning from contrastive learning to masked image modeling (MIM), there is no significant difference in essence, that is, how to design proper pretext tasks for vision dictionary look-up. MIM recently dominates this line of research with state-of-the-art performance on vision Transformers (ViTs), where the core is to enhance the patch-level visual context capturing of the network via denoising auto-encoding mechanism. Rather than tailoring image tokenizers with extra training stages as in previous works, we unleash the great potential of contrastive learning on denoising auto-encoding and introduce a pure MIM method, ConMIM, to produce simple intra-image inter-patch contrastive constraints as the sole learning objectives for masked patch prediction. We further strengthen the denoising mechanism with asymmetric designs, including image perturbations and model progress rates, to improve the network pre-training. ConMIM-pretrained models with various scales achieve competitive results on downstream image classification, semantic segmentation, object detection, and instance segmentation tasks, e.g., on ImageNet-1K classification, we achieve 83.9% top-1 accuracy with ViT-Small and 85.3% with ViT-Base without extra data for pre-training., Comment: Accepted by ICLR 2023. The code will be available at https://github.com/TencentARC/ConMIM

This paper focuses on self-supervised video representation learning. Most existing approaches follow the contrastive learning pipeline to construct positive and negative pairs by sampling different clips. However, this formulation tends to bias to static background and have difficulty establishing global temporal structures. The major reason is that the positive pairs, i.e., different clips sampled from the same video, have limited temporal receptive field, and usually share similar background but differ in motions. To address these problems, we propose a framework to jointly utilize local clips and global videos to learn from detailed region-level correspondence as well as general long-term temporal relations. Based on a set of controllable augmentations, we achieve accurate appearance and motion pattern alignment through soft spatio-temporal region contrast. Our formulation is able to avoid the low-level redundancy shortcut by mutual information minimization to improve the generalization. We also introduce local-global temporal order dependency to further bridge the gap between clip-level and video-level representations for robust temporal modeling. Extensive experiments demonstrate that our framework is superior on three video benchmarks in action recognition and video retrieval, capturing more accurate temporal dynamics.

In software engineering (SE) tasks, the naming approach is so important that it attracts many scholars from all over the world to study how to improve the quality of method names. To accurately recommend method names, we employ a novel framework to handle this problem. In our expeirments, nearly 8 million Java methods are collected from open source organizations as our evaluation dataset. In the first-phase recommendation, we introduce a fast and simple classifier based on the fast text neural network for reccomending potential method category. In the second-phase recomendation, we employ both two Long Short Term Memory Networks to reccomend consitent method names from each classification. Evaluation results prove that the proposed approach significantly outperforms state-of-the-art approach.

A large-scale evaluation for current naming approaches substantiates that such approaches are accurate. However, it is less known about which categories of method names work well via such naming approaches and how's the performance of naming approaches. To point out the superiority of the current naming approach, in this paper, we conduct an empirical study on such approaches in a new dataset. Moreover, we analyze the successful naming approaches above and find that: (1) around 60% of the accepted recommendation names are made on prefixes within get, set, is, and test. (2) A large portion (19.3%) of method names successfully recommended could be derived from the given method bodies. The comparisons also demonstrate the superior performance of the empirical study.

Pre-training a model to learn transferable video-text representation for retrieval has attracted a lot of attention in recent years. Previous dominant works mainly adopt two separate encoders for efficient retrieval, but ignore local associations between videos and texts. Another line of research uses a joint encoder to interact video with texts, but results in low efficiency since each text-video pair needs to be fed into the model. In this work, we enable fine-grained video-text interactions while maintaining high efficiency for retrieval via a novel pretext task, dubbed as Multiple Choice Questions (MCQ), where a parametric module BridgeFormer is trained to answer the "questions" constructed by the text features via resorting to the video features. Specifically, we exploit the rich semantics of text (i.e., nouns and verbs) to build questions, with which the video encoder can be trained to capture more regional content and temporal dynamics. In the form of questions and answers, the semantic associations between local video-text features can be properly established. BridgeFormer is able to be removed for downstream retrieval, rendering an efficient and flexible model with only two encoders. Our method outperforms state-of-the-art methods on the popular text-to-video retrieval task in five datasets with different experimental setups (i.e., zero-shot and fine-tune), including HowTo100M (one million videos). We further conduct zero-shot action recognition, which can be cast as video-to-text retrieval, and our approach also significantly surpasses its counterparts. As an additional benefit, our method achieves competitive results with much shorter pre-training videos on single-modality downstream tasks, e.g., action recognition with linear evaluation., Comment: Accepted by CVPR 2022