{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/ediffi-text-to-image-diffusion-models-with-an","title":"eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert Denoisers","arxiv_id":"2211.01324","date":"2022-11-02","proceeding":null,"authors":["Yogesh Balaji","Seungjun Nah","Xun Huang","Arash Vahdat","Jiaming Song","Qinsheng Zhang","Karsten Kreis","Miika Aittala","Timo Aila","Samuli Laine","Bryan Catanzaro","Tero Karras","Ming-Yu Liu"],"abstract":"Large-scale diffusion-based generative models have led to breakthroughs in text-conditioned high-resolution image synthesis. Starting from random noise, such text-to-image diffusion models gradually synthesize images in an iterative fashion while conditioning on text prompts. We find that their synthesis behavior qualitatively changes throughout this process: Early in sampling, generation strongly relies on the text prompt to generate text-aligned content, while later, the text conditioning is almost entirely ignored. This suggests that sharing model parameters throughout the entire generation process may not be ideal. Therefore, in contrast to existing works, we propose to train an ensemble of text-to-image diffusion models specialized for different synthesis stages. To maintain training efficiency, we initially train a single model, which is then split into specialized models that are trained for the specific stages of the iterative generation process. Our ensemble of diffusion models, called eDiff-I, results in improved text alignment while maintaining the same inference computation cost and preserving high visual quality, outperforming previous large-scale text-to-image diffusion models on the standard benchmark. In addition, we train our model to exploit a variety of embeddings for conditioning, including the T5 text, CLIP text, and CLIP image embeddings. We show that these different embeddings lead to different behaviors. Notably, the CLIP image embedding allows an intuitive way of transferring the style of a reference image to the target text-to-image output. Lastly, we show a technique that enables eDiff-I's \"paint-with-words\" capability. A user can select the word in the input text and paint it in a canvas to control the output, which is very handy for crafting the desired image in mind. The project page is available at https://deepimagination.cc/eDiff-I/","url_abs":"https://arxiv.org/abs/2211.01324v5","url_pdf":"https://arxiv.org/pdf/2211.01324v5.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"ediffi-text-to-image-diffusion-models-with-an","repo_url":"https://github.com/cloneofsimo/paint-with-words-sd","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"ediffi-text-to-image-diffusion-models-with-an","repo_url":"https://github.com/cross-domain-compositing/cross-domain-compositing","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"image-generation","task_name":"Image Generation"},{"task_slug":"text-to-image-generation","task_name":"Text-to-Image Generation"}],"methods":[{"method_slug":"adafactor","method_name":"Adafactor"},{"method_slug":"attention","method_name":"Attention"},{"method_slug":"attention-dropout","method_name":"Attention Dropout"},{"method_slug":"bpe","method_name":"BPE"},{"method_slug":"clip","method_name":"CLIP"},{"method_slug":"dense-connections","method_name":"Dense Connections"},{"method_slug":"diffusion","method_name":"Diffusion"},{"method_slug":"dropout","method_name":"Dropout"},{"method_slug":"glu","method_name":"Gated Linear Unit"},{"method_slug":"inverse-square-root-schedule","method_name":"Inverse Square Root Schedule"},{"method_slug":"layer-normalization","method_name":"Layer Normalization"},{"method_slug":"linear-layer","method_name":"Linear Layer"},{"method_slug":"multi-head-attention","method_name":"Multi-Head Attention"},{"method_slug":"residual-connection","method_name":"Residual Connection"},{"method_slug":"sentencepiece","method_name":"SentencePiece"},{"method_slug":"softmax","method_name":"Softmax"},{"method_slug":"t5","method_name":"T5"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/text-to-image-generation-on-coco","task":"Text-to-Image Generation","dataset":"COCO (Common Objects in Context)","model":"eDiff-I (zero-shot)","rank_in_archive_order":13,"of":69,"metrics":{"FID":"6.95"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2211.01324","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2211.01324"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/cross-domain-compositing/cross-domain-compositing","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/cloneofsimo/paint-with-words-sd","reach":null}],"summary":{"ran_honours":1,"ran":2,"ran_fixture":1,"unverified":9},"by_repo_kind":{"listed":{"samples":13,"ran":4,"repositories":2}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"a611fe4aef89ea52","entry":"always_round","repo":"cloneofsimo/paint-with-words-sd","repo_kind":"listed","path":"paint_with_words/paint_with_words.py","file_url":"https://github.com/cloneofsimo/paint-with-words-sd/blob/HEAD/paint_with_words/paint_with_words.py","link_basis":"first_harvest_node","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"well_formed","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"a611fe4aef89ea52"}},{"code_sha256_prefix":"314afb570b35e5c0","entry":"get_schedule","repo":"cross-domain-compositing/cross-domain-compositing","repo_kind":"listed","path":"cdc/scheduler.py","file_url":"https://github.com/cross-domain-compositing/cross-domain-compositing/blob/HEAD/cdc/scheduler.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"314afb570b35e5c0"}},{"code_sha256_prefix":"de7114b637a2e1b6","entry":"inj_forward","repo":"cloneofsimo/paint-with-words-sd","repo_kind":"listed","path":"paint_with_words/paint_with_words.py","file_url":"https://github.com/cloneofsimo/paint-with-words-sd/blob/HEAD/paint_with_words/paint_with_words.py","link_basis":"first_harvest_node","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"invariant","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"de7114b637a2e1b6"}},{"code_sha256_prefix":"cc073aabc2f49dab","entry":"preprocess","repo":"cloneofsimo/paint-with-words-sd","repo_kind":"listed","path":"paint_with_words/paint_with_words.py","file_url":"https://github.com/cloneofsimo/paint-with-words-sd/blob/HEAD/paint_with_words/paint_with_words.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":"well_formed","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"cc073aabc2f49dab"}},{"code_sha256_prefix":"99b6da8de85fe2bc","entry":"cross_attention_masks","repo":"cross-domain-compositing/cross-domain-compositing","repo_kind":"listed","path":"cdc/prompt_masking.py","file_url":"https://github.com/cross-domain-compositing/cross-domain-compositing/blob/HEAD/cdc/prompt_masking.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"99b6da8de85fe2bc"}},{"code_sha256_prefix":"4533e43e06a12dc5","entry":"download_models","repo":"cross-domain-compositing/cross-domain-compositing","repo_kind":"listed","path":"notebook_helpers.py","file_url":"https://github.com/cross-domain-compositing/cross-domain-compositing/blob/HEAD/notebook_helpers.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"4533e43e06a12dc5"}},{"code_sha256_prefix":"0657ed0cbf2c4fc5","entry":"get_blend_operator_and_mask","repo":"cross-domain-compositing/cross-domain-compositing","repo_kind":"listed","path":"cdc/cdc.py","file_url":"https://github.com/cross-domain-compositing/cross-domain-compositing/blob/HEAD/cdc/cdc.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"0657ed0cbf2c4fc5"}},{"code_sha256_prefix":"aefaff0f568e5037","entry":"get_schedule_jump","repo":"cross-domain-compositing/cross-domain-compositing","repo_kind":"listed","path":"cdc/scheduler.py","file_url":"https://github.com/cross-domain-compositing/cross-domain-compositing/blob/HEAD/cdc/scheduler.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"aefaff0f568e5037"}},{"code_sha256_prefix":"2032ab3924e8cd43","entry":"get_word_inds","repo":"cross-domain-compositing/cross-domain-compositing","repo_kind":"listed","path":"cdc/prompt_masking.py","file_url":"https://github.com/cross-domain-compositing/cross-domain-compositing/blob/HEAD/cdc/prompt_masking.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"2032ab3924e8cd43"}},{"code_sha256_prefix":"619771c3e74f395b","entry":"load_checkpoint","repo":"cross-domain-compositing/cross-domain-compositing","repo_kind":"listed","path":"SVR/utils.py","file_url":"https://github.com/cross-domain-compositing/cross-domain-compositing/blob/HEAD/SVR/utils.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"619771c3e74f395b"}},{"code_sha256_prefix":"f04426f214fbb1ec","entry":"load_model","repo":"cross-domain-compositing/cross-domain-compositing","repo_kind":"listed","path":"SVR/utils.py","file_url":"https://github.com/cross-domain-compositing/cross-domain-compositing/blob/HEAD/SVR/utils.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"f04426f214fbb1ec"}},{"code_sha256_prefix":"ce5593d971584146","entry":"process_image","repo":"cross-domain-compositing/cross-domain-compositing","repo_kind":"listed","path":"SVR/augment_ShapeNet_data.py","file_url":"https://github.com/cross-domain-compositing/cross-domain-compositing/blob/HEAD/SVR/augment_ShapeNet_data.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"ce5593d971584146"}},{"code_sha256_prefix":"08767bfaed57712a","entry":"sample_grid_points","repo":"cross-domain-compositing/cross-domain-compositing","repo_kind":"listed","path":"SVR/utils.py","file_url":"https://github.com/cross-domain-compositing/cross-domain-compositing/blob/HEAD/SVR/utils.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"08767bfaed57712a"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}