Multi-modal Large language models (MLLMs) show remarkable ability in video understanding. Nevertheless, understanding long videos remains challenging as the models can only process a finite number of frames in a single inference, potentially omitting crucial visual information. To address the challenge, we propose generating multiple predictions through visual context sampling, followed by a scoring mechanism to select the final prediction. Specifically, we devise a bin-wise sampling strategy that enables MLLMs to generate diverse answers based on various combinations of keyframes, thereby enriching the visual context. To determine the final prediction from the sampled answers, we employ a self-reward by linearly combining three scores: (1) a frequency score indicating the prevalence of each option, (2) a marginal confidence score reflecting the inter-intra sample certainty of MLLM predictions, and (3) a reasoning score for different question types, including clue-guided answering for global questions and temporal self-refocusing for local questions. The frequency score ensures robustness through majority correctness, the confidence-aligned score reflects prediction certainty, and the typed-reasoning score addresses cases with sparse key visual information using tailored strategies. Experiments show that this approach covers the correct answer for a high percentage of long video questions, on seven datasets show that our method improves the performance of three MLLMs.

Visual instructions for long-horizon tasks are crucial as they intuitively clarify complex concepts and enhance retention across extended steps. Directly generating a series of images using text-to-image models without considering the context of previous steps results in inconsistent images, increasing cognitive load. Additionally, the generated images often miss objects or the attributes such as color, shape, and state of the objects are inaccurate. To address these challenges, we propose LIGER, the first training-free framework for Long-horizon Instruction GEneration with logic and attribute self-Reflection. LIGER first generates a draft image for each step with the historical prompt and visual memory of previous steps. This step-by-step generation approach maintains consistency between images in long-horizon tasks. Moreover, LIGER utilizes various image editing tools to rectify errors including wrong attributes, logic errors, object redundancy, and identity inconsistency in the draft images. Through this self-reflection mechanism, LIGER improves the logic and object attribute correctness of the images. To verify whether the generated images assist human understanding, we manually curated a new benchmark consisting of various long-horizon tasks. Human-annotated ground truth expressions reflect the human-defined criteria for how an image should appear to be illustrative. Experiments demonstrate the visual instructions generated by LIGER are more comprehensive compared with baseline methods. The code and dataset will be available once accepted.

We study the zero-shot Composed Image Retrieval (ZS- CIR) task, which is to retrieve the target image given a reference image and a description without training on the triplet datasets. Previous works generate pseudo-word tokens by projecting the reference image features to the text embedding space. However, they focus on the global visual representation, ignoring the representation of detailed attributes, e.g., color, object number and layout. To address this challenge, we propose a Knowledge-Enhanced Dual-stream zero-shot composed image retrieval framework (KEDs). KEDs implicitly models the attributes of the reference images by incorporating a database. The database enriches the pseudo-word tokens by providing relevant images and captions, emphasizing shared attribute information in various aspects. In this way, KEDs recognizes the reference image from diverse perspectives. Moreover, KEDs adopts an extra stream that aligns pseudo-word tokens with textual concepts, leveraging pseudo-triplets mined from image-text pairs. The pseudo-word tokens generated in this stream are explicitly aligned with fine-grained semantics in the text embedding space. Extensive experiments on widely used benchmarks, i.e. ImageNet-R, COCO object, Fashion-IQ and CIRR, show that KEDs outperforms previous zero-shot composed image retrieval methods. Code is available at https://github.com/suoych/KEDs.

Sign language provides a way for differently-abled individuals to express their feelings and emotions. However, learning sign language can be challenging and time consuming. An alternative approach is to animate user photos using sign language videos of specific words, which can be achieved using existing image animation methods. However, the finger motions in the generated videos are often not ideal. To address this issue, we propose the Structure-aware Temporal Consistency Network (STCNet), which jointly optimizes the prior structure of humans with temporal consistency to produce sign language videos. We use a fine-grained skeleton detector to acquire knowledge of body structure and introduce both short- and long-term cycle loss to ensure the continuity of the generated video. The two losses and keypoint detector network are optimized in an end-to-end manner. Quantitative and qualitative evaluations on three widely used datasets, namely LSA64, Phoenix-2014T, and WLASL-2000, demonstrate the effectiveness of the proposed method. It is our hope that this work can contribute to future studies on sign language production.