Search

Dense prediction is a critical task in computer vision. However, previous methods often require extensive computational resources, which hinders their real-world application. In this paper, we propose BiDense, a generalized binary neural network (BNN) designed for efficient and accurate dense prediction tasks. BiDense incorporates two key techniques: the Distribution-adaptive Binarizer (DAB) and the Channel-adaptive Full-precision Bypass (CFB). The DAB adaptively calculates thresholds and scaling factors for binarization, effectively retaining more information within BNNs. Meanwhile, the CFB facilitates full-precision bypassing for binary convolutional layers undergoing various channel size transformations, which enhances the propagation of real-valued signals and minimizes information loss. By leveraging these techniques, BiDense preserves more real-valued information, enabling more accurate and detailed dense predictions in BNNs. Extensive experiments demonstrate that our framework achieves performance levels comparable to full-precision models while significantly reducing memory usage and computational costs.

State Space Models (SSM), such as Mamba, have shown strong representation ability in modeling long-range dependency with linear complexity, achieving successful applications from high-level to low-level vision tasks. However, SSM's sequential nature necessitates multiple scans in different directions to compensate for the loss of spatial dependency when unfolding the image into a 1D sequence. This multi-direction scanning strategy significantly increases the computation overhead and is unbearable for high-resolution image processing. To address this problem, we propose a novel Hierarchical Mamba network, namely, Hi-Mamba, for image super-resolution (SR). Hi-Mamba consists of two key designs: (1) The Hierarchical Mamba Block (HMB) assembled by a Local SSM (L-SSM) and a Region SSM (R-SSM) both with the single-direction scanning, aggregates multi-scale representations to enhance the context modeling ability. (2) The Direction Alternation Hierarchical Mamba Group (DA-HMG) allocates the isomeric single-direction scanning into cascading HMBs to enrich the spatial relationship modeling. Extensive experiments demonstrate the superiority of Hi-Mamba across five benchmark datasets for efficient SR. For example, Hi-Mamba achieves a significant PSNR improvement of 0.29 dB on Manga109 for $\times3$ SR, compared to the strong lightweight MambaIR.

Diffusion models have been achieving excellent performance for real-world image super-resolution (Real-ISR) with considerable computational costs. Current approaches are trying to derive one-step diffusion models from multi-step counterparts through knowledge distillation. However, these methods incur substantial training costs and may constrain the performance of the student model by the teacher's limitations. To tackle these issues, we propose DFOSD, a Distillation-Free One-Step Diffusion model. Specifically, we propose a noise-aware discriminator (NAD) to participate in adversarial training, further enhancing the authenticity of the generated content. Additionally, we improve the perceptual loss with edge-aware DISTS (EA-DISTS) to enhance the model's ability to generate fine details. Our experiments demonstrate that, compared with previous diffusion-based methods requiring dozens or even hundreds of steps, our DFOSD attains comparable or even superior results in both quantitative metrics and qualitative evaluations. Our DFOSD also abtains higher performance and efficiency compared with other one-step diffusion methods. We will release code and models at https://github.com/JianzeLi-114/DFOSD.

Knowledge distillation aims to transfer knowledge from a large teacher model to a compact student counterpart, often coming with a significant performance gap between them. We find that a too-large performance gap can hamper the training process, which is also verified in recent studies. To address this, we propose a Gap Preserving Distillation (GPD) method that trains an additional dynamic teacher model from scratch along with training the student to bridge this gap. In this way, it becomes possible to maintain a reasonable performance gap between teacher and student during the whole distillation process. To further strengthen distillation from the dynamic teacher to the student, we develop a hard strategy by enforcing them to share parameters and encouraging parameter inheritance. Besides hard strategy, we also build the soft bidirectional mappings between them which are built on an Inverse Reparameterization (IR) method and a Channel-Branch Reparameterization (CBR) strategy. We highlight that our IR is able to initialize a larger dynamic teacher with an arbitrary expansion ratio, while preserving exactly the same accuracy as the given student model. In this way, it guarantees that the dynamic teacher and student start from the same point and avoid a too large gap in early stage of training. As for our CBR, with parameter-sharing, it directly extracts an effective student model from the well-learned dynamic teacher without any post-training, making our method highly flexible for model deployment. In the experiments, GPD significantly outperforms existing distillation methods on top of both CNNs and transformers architectures, achieving up to 1.58% accuracy improvement. Interestingly, GPD also generalizes well to the scenarios without a pre-trained teacher, including training from scratch and fine-tuning, yielding a large improvement of 1.80% and 0.89% on ResNet18, respectively., Comment: 10 pages for the main paper

Recent advances indicate that diffusion models hold great promise in image super-resolution. While the latest methods are primarily based on latent diffusion models with convolutional neural networks, there are few attempts to explore transformers, which have demonstrated remarkable performance in image generation. In this work, we design an effective diffusion transformer for image super-resolution (DiT-SR) that achieves the visual quality of prior-based methods, but through a training-from-scratch manner. In practice, DiT-SR leverages an overall U-shaped architecture, and adopts a uniform isotropic design for all the transformer blocks across different stages. The former facilitates multi-scale hierarchical feature extraction, while the latter reallocates the computational resources to critical layers to further enhance performance. Moreover, we thoroughly analyze the limitation of the widely used AdaLN, and present a frequency-adaptive time-step conditioning module, enhancing the model's capacity to process distinct frequency information at different time steps. Extensive experiments demonstrate that DiT-SR outperforms the existing training-from-scratch diffusion-based SR methods significantly, and even beats some of the prior-based methods on pretrained Stable Diffusion, proving the superiority of diffusion transformer in image super-resolution., Comment: Code is available at https://github.com/kunncheng/DiT-SR

Diffusion-based image super-resolution (SR) methods have shown promise in reconstructing high-resolution images with fine details from low-resolution counterparts. However, these approaches typically require tens or even hundreds of iterative samplings, resulting in significant latency. Recently, techniques have been devised to enhance the sampling efficiency of diffusion-based SR models via knowledge distillation. Nonetheless, when aligning the knowledge of student and teacher models, these solutions either solely rely on pixel-level loss constraints or neglect the fact that diffusion models prioritize varying levels of information at different time steps. To accomplish effective and efficient image super-resolution, we propose a time-aware diffusion distillation method, named TAD-SR. Specifically, we introduce a novel score distillation strategy to align the data distribution between the outputs of the student and teacher models after minor noise perturbation. This distillation strategy enables the student network to concentrate more on the high-frequency details. Furthermore, to mitigate performance limitations stemming from distillation, we integrate a latent adversarial loss and devise a time-aware discriminator that leverages diffusion priors to effectively distinguish between real images and generated images. Extensive experiments conducted on synthetic and real-world datasets demonstrate that the proposed method achieves comparable or even superior performance compared to both previous state-of-the-art (SOTA) methods and the teacher model in just one sampling step. Codes are available at https://github.com/LearningHx/TAD-SR., Comment: 18 pages

Radiofrequency harvesting using ambient wireless energy could be used to reduce the carbon footprint of electronic devices. However, ambient radiofrequency energy is weak (less than -20 dBm), and thermodynamic limits and high-frequency parasitic impedance restrict the performance of state-of-the-art radiofrequency rectifiers. Nanoscale spin rectifiers based on magnetic tunnel junctions have recently demonstrated high sensitivity, but suffer from a low a.c.-to-d.c. conversion efficiency (less than 1%). Here, we report a sensitive spin rectifier rectenna that can harvest ambient radiofrequency signals between -62 and -20 dBm. We also develop an on-chip co-planar waveguide-based spin rectifier array with a large zero-bias sensitivity (around 34,500 mV/mW) and high efficiency (7.81%). Self-parametric excitation driven by voltage-controlled magnetic anisotropy is a key mechanism that contributes to the performance of the spin-rectifier array. We show that these spin rectifiers can wirelessly power a sensor at a radiofrequency power of -27 dBm.

Real-world data often follows a long-tailed distribution, where a few head classes occupy most of the data and a large number of tail classes only contain very limited samples. In practice, deep models often show poor generalization performance on tail classes due to the imbalanced distribution. To tackle this, data augmentation has become an effective way by synthesizing new samples for tail classes. Among them, one popular way is to use CutMix that explicitly mixups the images of tail classes and the others, while constructing the labels according to the ratio of areas cropped from two images. However, the area-based labels entirely ignore the inherent semantic information of the augmented samples, often leading to misleading training signals. To address this issue, we propose a Contrastive CutMix (ConCutMix) that constructs augmented samples with semantically consistent labels to boost the performance of long-tailed recognition. Specifically, we compute the similarities between samples in the semantic space learned by contrastive learning, and use them to rectify the area-based labels. Experiments show that our ConCutMix significantly improves the accuracy on tail classes as well as the overall performance. For example, based on ResNeXt-50, we improve the overall accuracy on ImageNet-LT by 3.0% thanks to the significant improvement of 3.3% on tail classes. We highlight that the improvement also generalizes well to other benchmarks and models. Our code and pretrained models are available at https://github.com/PanHaulin/ConCutMix., Comment: 16 pages and 13 figures

Ultra-high-resolution image generation poses great challenges, such as increased semantic planning complexity and detail synthesis difficulties, alongside substantial training resource demands. We present UltraPixel, a novel architecture utilizing cascade diffusion models to generate high-quality images at multiple resolutions (\textit{e.g.}, 1K to 6K) within a single model, while maintaining computational efficiency. UltraPixel leverages semantics-rich representations of lower-resolution images in the later denoising stage to guide the whole generation of highly detailed high-resolution images, significantly reducing complexity. Furthermore, we introduce implicit neural representations for continuous upsampling and scale-aware normalization layers adaptable to various resolutions. Notably, both low- and high-resolution processes are performed in the most compact space, sharing the majority of parameters with less than 3$\%$ additional parameters for high-resolution outputs, largely enhancing training and inference efficiency. Our model achieves fast training with reduced data requirements, producing photo-realistic high-resolution images and demonstrating state-of-the-art performance in extensive experiments., Comment: Project page https://jingjingrenabc.github.io/ultrapixel

Low-bit quantization has become widespread for compressing image super-resolution (SR) models for edge deployment, which allows advanced SR models to enjoy compact low-bit parameters and efficient integer/bitwise constructions for storage compression and inference acceleration, respectively. However, it is notorious that low-bit quantization degrades the accuracy of SR models compared to their full-precision (FP) counterparts. Despite several efforts to alleviate the degradation, the transformer-based SR model still suffers severe degradation due to its distinctive activation distribution. In this work, we present a dual-stage low-bit post-training quantization (PTQ) method for image super-resolution, namely 2DQuant, which achieves efficient and accurate SR under low-bit quantization. The proposed method first investigates the weight and activation and finds that the distribution is characterized by coexisting symmetry and asymmetry, long tails. Specifically, we propose Distribution-Oriented Bound Initialization (DOBI), using different searching strategies to search a coarse bound for quantizers. To obtain refined quantizer parameters, we further propose Distillation Quantization Calibration (DQC), which employs a distillation approach to make the quantized model learn from its FP counterpart. Through extensive experiments on different bits and scaling factors, the performance of DOBI can reach the state-of-the-art (SOTA) while after stage two, our method surpasses existing PTQ in both metrics and visual effects. 2DQuant gains an increase in PSNR as high as 4.52dB on Set5 (x2) compared with SOTA when quantized to 2-bit and enjoys a 3.60x compression ratio and 5.08x speedup ratio. The code and models will be available at https://github.com/Kai-Liu001/2DQuant., Comment: 9 pages, 6 figures. The code and models will be available at https://github.com/Kai-Liu001/2DQuant

Advanced diffusion models (DMs) perform impressively in image super-resolution (SR), but the high memory and computational costs hinder their deployment. Binarization, an ultra-compression algorithm, offers the potential for effectively accelerating DMs. Nonetheless, due to the model structure and the multi-step iterative attribute of DMs, existing binarization methods result in significant performance degradation. In this paper, we introduce a novel binarized diffusion model, BI-DiffSR, for image SR. First, for the model structure, we design a UNet architecture optimized for binarization. We propose the consistent-pixel-downsample (CP-Down) and consistent-pixel-upsample (CP-Up) to maintain dimension consistent and facilitate the full-precision information transfer. Meanwhile, we design the channel-shuffle-fusion (CS-Fusion) to enhance feature fusion in skip connection. Second, for the activation difference across timestep, we design the timestep-aware redistribution (TaR) and activation function (TaA). The TaR and TaA dynamically adjust the distribution of activations based on different timesteps, improving the flexibility and representation alability of the binarized module. Comprehensive experiments demonstrate that our BI-DiffSR outperforms existing binarization methods. Code is released at: https://github.com/zhengchen1999/BI-DiffSR., Comment: Accepted to NeurIPS 2024. Code is available at https://github.com/zhengchen1999/BI-DiffSR

Single-photon sources are essential for quantum networks, enabling applications ranging from quantum key distribution (QKD) to the burgeoning quantum internet. Despite the remarkable advancements, the current reliance of QKD on attenuated coherent (laser) light sources has imposed a fundamental limit on the secret key rate (SKR). This constraint is primarily attributable to the scarcity of single-photon components within coherent light, confined by an inherent upper bound of 1/e. Here, we report high-rate QKD using a high-efficiency single-photon source, enabling an SKR transcending the fundamental rate limit of coherent light. We developed an on-demand, bright semiconductor quantum-dot single-photon source with an efficiency of 0.71(2), exceeding the inherent bound of coherent light by approximately 2.87 dB. Implementing narrow-bandwidth filtering and random polarization modulation, we conducted a field QKD trial over a 14.6(1.1)-dB-loss free-space urban channel, achieving an SKR of 0.00108 bits per pulse. This surpasses the practical limit of coherent-light-based QKD by 2.53 dB. Our findings conclusively demonstrate the superior performance of nanotechnology-based single-photon sources over coherent light for QKD applications, marking a pivotal stride towards the realization of a global quantum internet., Comment: 22 pages, 5 figures, 1 Table

Temporal action detection (TAD) aims to locate action positions and recognize action categories in long-term untrimmed videos. Although many methods have achieved promising results, their robustness has not been thoroughly studied. In practice, we observe that temporal information in videos can be occasionally corrupted, such as missing or blurred frames. Interestingly, existing methods often incur a significant performance drop even if only one frame is affected. To formally evaluate the robustness, we establish two temporal corruption robustness benchmarks, namely THUMOS14-C and ActivityNet-v1.3-C. In this paper, we extensively analyze the robustness of seven leading TAD methods and obtain some interesting findings: 1) Existing methods are particularly vulnerable to temporal corruptions, and end-to-end methods are often more susceptible than those with a pre-trained feature extractor; 2) Vulnerability mainly comes from localization error rather than classification error; 3) When corruptions occur in the middle of an action instance, TAD models tend to yield the largest performance drop. Besides building a benchmark, we further develop a simple but effective robust training method to defend against temporal corruptions, through the FrameDrop augmentation and Temporal-Robust Consistency loss. Remarkably, our approach not only improves robustness but also yields promising improvements on clean data. We believe that this study will serve as a benchmark for future research in robust video analysis. Source code and models are available at https://github.com/Alvin-Zeng/temporal-robustness-benchmark., Comment: Accepted by CVPR2024

Dissipative Rydberg gases are an outstanding platform for the investigation of many-body quantum open systems. Despite the wealth of existing studies, the non-equilibrium dynamics of dissipative Rydberg gases are rarely examined or harnessed from the perspective of non-Hermitian physics, which is but intrinsic to open systems. Here we report the experimental observation of a chiral switching between many-body steady states in a dissipative thermal Rydberg vapor, where the interplay of many-body effects and non-Hermiticity plays a key role. Specifically, as the parameters are adiabatically varied around a closed contour, depending on the chirality of the parameter modulation, the Rydberg vapor can change between two collective steady states with distinct Rydberg excitations and optical transmissions. Adopting a mean-field description, we reveal that both the existence of the bistable steady states and chiral dynamics derive from an exceptional structure in the parameter space, where multiple steady states of the many-body Liouvillian superoperator coalesce. We demonstrate that both the exceptional structure and the resulting state-switching dynamics are tunable through microwave dressing and temperature variations, confirming their reliance on the many-body dissipative nature of the Rydberg vapor.

In practical sensing tasks, noise is usually regarded as an obstruction to better performance and will degrade the sensitivity. Fortunately, \textit{stochastic resonance} (SR), a counterintuitive concept, can utilize noise to greatly enhance the detected signal amplitude. Although fundamentally important as a mechanism of weak signal amplification, and has been continually explored in geological, biological, and physical science, both theoretically and experimentally, SR has yet to be demonstrated in realistic sensing tasks. Here we develop a novel SR-based nonlinear sensor using a thermal ensemble of interacting Rydberg atoms. With the assistance of stochastic noise (either inherently in the system or added externally) and strong nonlinearity in the Rydberg ensembles, the signal encoded in a weak MW field is greatly enhanced (over 25 dB). Moreover, we show that the SR-based atomic sensor can achieve a better sensitivity in our system, which is over 6.6 dB compared to a heterodye atomic sensor. Our results show potential advantage of SR-based MW sensors in commercial or defense-related applications.

Customized text-to-video generation aims to generate high-quality videos guided by text prompts and subject references. Current approaches for personalizing text-to-video generation suffer from tackling multiple subjects, which is a more challenging and practical scenario. In this work, our aim is to promote multi-subject guided text-to-video customization. We propose CustomVideo, a novel framework that can generate identity-preserving videos with the guidance of multiple subjects. To be specific, firstly, we encourage the co-occurrence of multiple subjects via composing them in a single image. Further, upon a basic text-to-video diffusion model, we design a simple yet effective attention control strategy to disentangle different subjects in the latent space of diffusion model. Moreover, to help the model focus on the specific area of the object, we segment the object from given reference images and provide a corresponding object mask for attention learning. Also, we collect a multi-subject text-to-video generation dataset as a comprehensive benchmark, with 63 individual subjects from 13 different categories and 68 meaningful pairs. Extensive qualitative, quantitative, and user study results demonstrate the superiority of our method compared to previous state-of-the-art approaches. The project page is https://kyfafyd.wang/projects/customvideo., Comment: 18 pages, 11 figures, 7 tables

Nonclassical phenomena tied to entangled states are focuses of foundational studies and powerful resources in many applications. By contrast, the counterparts on quantum measurements are still poorly understood. Notably, genuine multipartite nonclassicality is barely discussed, not to say experimental realization. Here we experimentally demonstrate the power of genuine tripartite nonclassicality in quantum measurements based on a simple estimation problem. To this end we realize an optimal genuine three-copy collective measurement via a nine-step two-dimensional photonic quantum walk with 30 elaborately designed coin operators. Then we realize an optimal estimation protocol and achieve an unprecedented high estimation fidelity, which can beat all strategies based on restricted collective measurements by more than 11 standard deviations. These results clearly demonstrate that genuine collective measurements can extract more information than local measurements and restricted collective measurements. Our work opens the door for exploring genuine multipartite nonclassical measurements and their power in quantum information processing.

Photon loss is the biggest enemy for scalable photonic quantum information processing. This problem can be tackled by using quantum error correction, provided that the overall photon loss is below a threshold of 1/3. However, all reported on-demand and indistinguishable single-photon sources still fall short of this threshold. Here, by using tailor shaped laser pulse excitation on a high-quantum efficiency single quantum dot deterministically coupled to a tunable open microcavity, we demonstrate a high-performance source with a single-photon purity of 0.9795(6), photon indistinguishability of 0.986(16), and an overall system efficiency of 0.717(20), simultaneously. This source for the first time reaches the efficiency threshold for scalable photonic quantum computing. With this source, we further demonstrate 1.87(13) dB intensity squeezing, and consecutive 40-photon events with 1.67 mHz count rate.

We investigate theoretically the enhancement of mechanical squeezing in a multimode optomechanical system by introducing a coherent phonon-photon interaction via the backward stimulated Brillouin scattering (BSBS) process. The coherent photon-phonon interaction where two optical modes couple to a Brillouin acoustic mode with a large decay rate provides an extra channel for the cooling of a Duffing mechanical oscillator. The squeezing degree and the robustness to the thermal noises of the Duffing mechanical mode can be enhanced greatly. When the Duffing nonlinearity is weak, the squeezing degree of the mechanical mode in the presence of BSBS can be improved more than one order of magnitude compared with the absence of BSBS. Our scheme may be extended to other quantum systems to study novel quantum effects.

Quantum measurements based on mutually unbiased bases (MUB) play crucial roles in foundational studies and quantum information processing. It is known that there exist inequivalent MUB, but little is known about their operational distinctions, not to say experimental demonstration. In this work, by virtue of a simple estimation problem we experimentally demonstrate the operational distinctions between inequivalent triples of MUB in dimension 4 based on high-precision photonic systems. The experimental estimation fidelities coincide well with the theoretical predictions with only 0.16$\%$ average deviation, which is 25 times less than the difference (4.1$\%$) between the maximum estimation fidelity and the minimum estimation fidelity. Our experiments clearly demonstrate that inequivalent MUB have different information extraction capabilities and different merits for quantum information processing.

Large language models (LLMs) have showcased remarkable prowess in code generation. However, automated code generation is still challenging since it requires a high-level semantic mapping between natural language requirements and codes. Most existing LLMs-based approaches for code generation rely on decoder-only causal language models often treate codes merely as plain text tokens, i.e., feeding the requirements as a prompt input, and outputing code as flat sequence of tokens, potentially missing the rich semantic features inherent in source code. To bridge this gap, this paper proposes the "Semantic Chain-of-Thought" approach to intruduce semantic information of code, named SeCoT. Our motivation is that the semantic information of the source code (\eg data flow and control flow) describes more precise program execution behavior, intention and function. By guiding LLM consider and integrate semantic information, we can achieve a more granular understanding and representation of code, enhancing code generation accuracy. Meanwhile, while traditional techniques leveraging such semantic information require complex static or dynamic code analysis to obtain features such as data flow and control flow, SeCoT demonstrates that this process can be fully automated via the intrinsic capabilities of LLMs (i.e., in-context learning), while being generalizable and applicable to challenging domains. While SeCoT can be applied with different LLMs, this paper focuses on the powerful GPT-style models: ChatGPT(close-source model) and WizardCoder(open-source model). The experimental study on three popular DL benchmarks (i.e., HumanEval, HumanEval-ET and MBPP) shows that SeCoT can achieves state-of-the-art performance, greatly improving the potential for large models and code generation., Comment: There may be calculation errors in Table 4 of the paper. We need time to verify and supplement, so the manuscript needs to be withdrawn. Thanks!

Multimodal large language models (MLLMs) have emerged as a prominent area of interest within the research community, given their proficiency in handling and reasoning with non-textual data, including images and videos. This study seeks to extend the application of MLLMs to the realm of autonomous driving by introducing DriveGPT4, a novel interpretable end-to-end autonomous driving system based on LLMs. Capable of processing multi-frame video inputs and textual queries, DriveGPT4 facilitates the interpretation of vehicle actions, offers pertinent reasoning, and effectively addresses a diverse range of questions posed by users. Furthermore, DriveGPT4 predicts low-level vehicle control signals in an end-to-end fashion.These advanced capabilities are achieved through the utilization of a bespoke visual instruction tuning dataset, specifically tailored for autonomous driving applications, in conjunction with a mix-finetuning training strategy. DriveGPT4 represents the pioneering effort to leverage LLMs for the development of an interpretable end-to-end autonomous driving solution. Evaluations conducted on the BDD-X dataset showcase the superior qualitative and quantitative performance of DriveGPT4. Additionally, the fine-tuning of domain-specific data enables DriveGPT4 to yield close or even improved results in terms of autonomous driving grounding when contrasted with GPT4-V., Comment: Accepted by RA-L. The project page is available at https://tonyxuqaq.github.io/projects/DriveGPT4/

Reducing the average resource consumption is the central quest in discriminating non-orthogonal quantum states for a fixed admissible error rate $\varepsilon$. The globally optimal fixed local projective measurement (GOFL) for this task is found to be different from that for previous minimum-error discrimination tasks [PRL 118, 030502 (2017)]. To achieve the ultimate minimum average consumption, here we develop a general globally optimal adaptive strategy (GOA) by subtly using the updated posterior probability, which works under any error rate requirement and any one-way measurement restrictions, and can be solved by a convergent iterative relation. First, under the local measurement restrictions, our GOA is solved to serve as the local bound, which saves 16.6 copies (24%) compared with the previously best GOFL. When the more powerful two-copy collective measurements are allowed, our GOA is experimentally demonstrated to beat the local bound by 3.9 copies (6.0%). By exploiting both adaptivity and collective measurements, our work marks an important step towards minimum-consumption quantum state discrimination.

In the quest to build general-purpose photonic quantum computers, fusion-based quantum computation has risen to prominence as a promising strategy. This model allows a ballistic construction of large cluster states which are universal for quantum computation, in a scalable and loss-tolerant way without feed-forward, by fusing many small n-photon entangled resource states. However, a key obstacle to this architecture lies in efficiently generating the required essential resource states on photonic chips. One such critical seed state that has not yet been achieved is the heralded three-photon Greenberger-Horne-Zeilinger (3-GHZ) state. Here, we address this elementary resource gap, by reporting the first experimental realization of a heralded dual-rail encoded 3-GHZ state. Our implementation employs a low-loss and fully programmable photonic chip that manipulates six indistinguishable single photons of wavelengths in the telecommunication regime. Conditional on the heralding detection, we obtain the desired 3-GHZ state with a fidelity 0.573+-0.024. Our work marks an important step for the future fault-tolerant photonic quantum computing, leading to the acceleration of building a large-scale optical quantum computer., Comment: Submitted. Comments are welcome!