NExT-Chat: An LMM for Chat, Detection and Segmentation
This work addresses the need for improved region-level understanding in multimodal AI systems, offering a flexible approach for tasks like detection and segmentation, but it appears incremental as it builds on existing pix2seq methods.
The paper tackles the problem of enhancing visual comprehension in large multimodal models by introducing a novel paradigm called pix2emb for object location modeling, which allows handling multiple tasks like visual grounding and region captioning, achieving results such as 87.7 vs. 86.9 on POPE-Random and 79.6 vs. 62.3 on region captioning.
The development of large language models (LLMs) has greatly advanced the field of multimodal understanding, leading to the emergence of large multimodal models (LMMs). In order to enhance the level of visual comprehension, recent studies have equipped LMMs with region-level understanding capabilities by representing object bounding box coordinates as a series of text sequences (pix2seq). In this paper, we introduce a novel paradigm for object location modeling called pix2emb method, where we ask the LMM to output the location embeddings and then decode them with different decoders. This paradigm allows us to use different location formats (such as bounding boxes and masks) in multimodal conversations. Leveraging the proposed pix2emb method, we train an LMM named NExT-Chat and demonstrate its capability of handling multiple tasks like visual grounding, region captioning, and grounded reasoning. Comprehensive experiments show the effectiveness of our NExT-Chat on various tasks, e.g., NExT-Chat (87.7) vs. Shikra (86.9) on POPE-Random, NExT-Chat (68.9) vs. LISA (67.9) on referring expression segmentation task, and NExT-Chat (79.6) vs. Kosmos-2 (62.3) on region caption task. The code and model are released at https://github.com/NExT-ChatV/NExT-Chat.